Structural Brain Differences between Men
and Women in Cocaine-users with
Childhood Trauma
AbstractIntroduction. Cocaine use disorder prevalence has been growing rapidly
over the last years. At the same time, sex differences in cocaine use are scarcely investigated. One of the factors contributing to sex differences in cocaine users could be childhood trauma (CT). In previous research it has been shown that CT is a predictor of cocaine substance use disorder later in life in women, but not in men. This study therefore aimed to investigate whether CT could partially explain sex differences in the neurobiology of cocaine use disorder. It is expected that the brain regions involved in cocaine use and CT are smaller in cocaine users compared to non-drug using controls, but in female cocaine users especially.
Methods. In this research, 57 (26 women) cocaine-using (CU) participants
and 56 (28 women) non-drug using controls (NC) were investigated. Questionnaires were used to assess childhood trauma and cocaine use. Grey matter morphometry was measured using a structural magnetic resonance imaging scan.
Results. In the assessment of determining childhood trauma, women had
significantly higher scores than men, and cocaine using women had
significantly higher scores than cocaine using men. Correcting for effects of childhood trauma on the brain, there was a main effect of sex,
exhibiting smaller grey matter volumes in women in the right caudate nucleus and the left putamen compared to men, and men showing a smaller left supplementary motor area (left SMA) than women. However, this correlation in the left SMA was mostly caused by a group by sex effect, which showed that CU men had a smaller volume in the left SMA compared to CU women. No main group effect was found. CT was shown to have a positive correlation with the left middle frontal cortex (left MFC), and a negative correlation with the left superior/middle temporal cortex. On top of that, a CT by sex effect was found, showing that women had a positive correlation with childhood trauma in the right middle cingulate cortex (right MCC), and men a negative correlation with childhood trauma in the right MCC.
Conclusion. What can be concluded from this study is that there are sex
differences, however these differences are mostly not dependent of cocaine use. CT showed to correlate with sex differences in the right MCC, however no other effects of childhood trauma nor cocaine use were found. This study makes it clear more research must be done about the sex differences, and how these affect cocaine use and grey matter
morphometry. More knowledge in sex differences could lead to more individualised treatment plans in the future.
Keywords: cocaine use, sex differences, childhood trauma,
structural MRI, prefrontal cortex, right middle frontal cortex, left supplementary motor area
Author: R. A. Hetterscheid
Studentnumber: 10725539 Supervisor: A.M. Kaag
Introduction
Substance use, and cocaine use especially, is an ever-growing concern in our society (Rhemtulla et al., 2016). According to the EMCDDA, there has been an increase of 25% in people seeking treatment for cocaine
substance use since 2014 (Europees Drugs rapport, 2019). Substance use disorder (SUD) is characterized by three stages (Becker & Koob, 2016): 1) Binge/intoxication: after initially taking the drug, more of the drug is needed to achieve the same “high”. Finding you have lost control over the amount you consume arises; 2) Withdrawal/negative affect: if the drug is not taken in time, withdrawal symptoms will arise and a negative
emotional state emerges when the drug is not consumed; 3)
Preoccupation/anticipation: Feeling compelled to find and take the drug, most of the times at cost of other events in someone’s life (Koob, 2009). Even though this is the general pattern of how SUD initiates itself, there are differences between men and women regarding the development of SUD and reacting to substances (Yang, Sau, Lai, Cichon, & Li, 2015). For example, while fewer women use cocaine compared to men, women tend to use more cocaine after the first exposure to the drug (Anker, J.J.,
Carroll, 2010). Moreover, women tend to escalate quicker into cocaine use disorder (CUD) than men (Bobzean, DeNobrega, & Perrotti, 2014). This has been shown in animal research as well: female rats have quicker escalation and relapse than male rats (Lynch, Roth, & Carroll, 2002). This has been replicated in different studies with different drugs of use
(Becker & Koob, 2016; Lynch et al., 2002). Despite the known
psychological differences between men and women, the neurobiological mechanisms that underlie these differences are largely unknown. It would be clinically relevant to gain more insight into the differences between men and women since treatment could be more fine-tuned to the
individual and be potentially more successful.
One of the mechanisms involved in the development of CUD is the correlation with smaller brain region volumes. Earlier research shows that CUD is linked with smaller grey matter volume in the anterior
cingulate cortex (ACC) (Beck et al., 2012; Casey & Jones, 2010; Franklin et al., 2002; Holmes, Hollinshead, Roffman, Smoller, & Buckner, 2016), middle frontal gyrus (Casey & Jones, 2010; Holmes et al., 2016),
amygdala (Makris et al., 2004), orbitofrontal cortex (OFC) (Beck et al., 2012; Franklin et al., 2002), insula (Franklin et al., 2002), superior temporal cortex (Franklin et al., 2002), prefrontal cortex (PFC) (Beck et al., 2012) and the hippocampus (HC) (Teicher, Anderson, & Polcari, 2012). Firstly, reduction in grey matter volume in the ACC, PFC and
middle frontal gyrus are linked to goal-directed attention and inhibition of impulsive actions, and secondly, reduction in grey matter volume in the PFC, OFC, and ACC is associated with a higher chance of relapse (Beck et al., 2012; Casey & Jones, 2010). Because of the known differences
between men and women regarding the development of CUD, it can be expected that differences between men and women can be seen in the above-mentioned brain regions (Becker & Koob, 2016), although this cannot be confirmed by research, as not a lot of research has been done
about sex differences in cocaine use. However, a possible explanation of sex differences in the development of CUD could be childhood trauma (CT). CT is a risk factor for SUD, including CUD (Enoch et al., 2010). The effects of CT can also be found in animals. In research of Kosten,
Sanchez, Zhang, & Kehoe (2004) it has been shown that CT1 is linked to
higher self-administration in rats. Interestingly, CT occurs more often in women than in men (Elliott, Mok, & Briere, 2004) and it has been shown that CT is associated with a higher risk for substance use (Widom, White, Czaja, & Marmorstein, 2007) and relapse (Hyman et al., 2009) in women but not in men. Unfortunately, the neurobiological mechanisms that
underlie these sex differences in the association between CT and CUD are largely unknown.
Animal studies demonstrated that CT is linked to a smaller HC volume in non-primate monkeys (Jackowski et al., 2011). On top of that, it has also been shown that CT is associated with decreased neurogenesis in the HC after puberty in rats, with the biggest reduction in female rats (Loi,
Koricka, Lucassen, & Joëls, 2014). Smaller HC volumes can also be found in humans. Gorka, Hanson, Radtke, & Hariri (2014) showed that CT associates with a smaller HC volume, and additionally a smaller medial PFC volume. On top of that, smaller volumes in the ACC, caudate nucleus, OFC and insula were found to be associated with CT (Farrell, Holland, Shansky, & Brenhouse, 2016; Saleh et al., 2017).
Even though a lot is known about the psychological associations of
cocaine use, sex and CT and how these three factors correlate with each other, unfortunately the exact relationship between CT, CUD, sex and grey matter morphometry has not been investigated yet.
That is why the aim of this research is to determine if sex differences in cocaine users is partially explained through CT. On basis of earlier
research, it will be predicted that CT partially explains the psychological sex differences in cocaine users, and since CT poses a great risk for cocaine use, it also partially explains the differences between cocaine users and non-drug using controls. On top of that, cocaine using women will show larger brain region abnormalities associated with cocaine use and CT in grey matter morphometry to their male counterparts, (Hyman et al., 2009).
The brain regions that will show a smaller volume include regions
associated with cocaine use in the general regions of the PFC, insula and superior temporal cortex (Beck et al., 2012; Casey & Jones, 2010;
Franklin et al., 2002; Holmes et al., 2016) compared to non-drug using controls, and on top of that regions associated with CT, which include: the HC, OFC, insula, PFC, ACC and the caudate nucleus (Cohen et al., 2006; Farrell et al., 2016; Gorka et al., 2014; Jackowski et al., 2011; Loi et al., 2014) compared to non-drug using controls. There will also be
investigated exploratively if the relation of sex and grey matter morphometry is influenced by other addiction-related variables, for instance the severity of addiction to cocaine, alcohol and cannabis.
1 In animal studies childhood trauma is generally called early life stress (ELS). However, in this research it will be called childhood trauma to make the study more cohesive.
Subjects and Methods Participants
In total there were 57 (26 women) regular cocaine users (CU) and 56 (28 women) non-drug using controls (NC) that were included in this research. Participants were recruited through online advertisements and poster advertisements in the city of Amsterdam. Before the participants were invited to the research centre, there were asked to complete an online screening to determine if they were fit to join the project and had to sign an online informed consent. Inclusion criteria for the regular cocaine user group were that they had to be snorting cocaine at least 4 times a month. They were excluded if they did not identify with either being a men or women, if someone had a background with severe neurological disease, if someone showed indications of posttraumatic stress disorder (PTSD) based on the primary care posttraumatic stress disorder screening
questionnaire (PC-PTSD) (van Dam, Ehring, Vedel, & Emmelkamp, 2010), if their age was below 18 or above 45 or if someone had
contra-indications for MRI scanning. On top of that, NCs were also excluded if they were regular smokers of nicotine, had used drugs more than 5 times in the past year or used cocaine more than 5 times in their life. NCs were also excluded if they had a score of 12 or higher on the Alcohol Use
Disorder Identification Test (AUDIT) (Lawford et al., 2012).
All participants signed an informed consent before testing and received monetary compensation for participating in the research. This research was approved by the Psychology Ethics Committee of the University of Amsterdam.
General procedure
For the online screening, the participants were asked to sign an online informed consent. After the online screening, the participants were invited to the research centre. When the participants arrived at the research centre, they were informed about the research and had to sign another informed consent before starting the testing. The participants were asked not to use drugs 24 hours prior to arriving to the research centre. The participants then had to complete the clinical and
demographic assessment and afterward had a structural magnetic resonance imaging scan (MRI-scan).
Clinical and demographic assessment
To determine the severity of childhood trauma(CT), the Dutch version of the Childhood Trauma Questionnaire (CTQ) was used (Bernstein et al., 2003). The total of the CTQ score gives an indication of the severity of CT. Participants were asked how much cocaine they used during the online screening and again at the research centre to determine cocaine use per month.2 In the screening questionnaire, the DSM-V (American Psychiatric
Association, 2013) was used to determine if they aligned with the diagnosis of cocaine use. The Drug Use Identification Test (DUDIT)
2 Initially, the timeline questionnaire was used to determine cocaine use per month since this is the most reliable test to determine cocaine use, but since a moderate amount of regular cocaine users did not complete this questionnaire, it was not used.
(Berman, Bergman, Palmstierna, & Schlyter, 2003) was used to determine severity of cocaine use, while the Desire for Drug
Questionnaire (DDQ) (Franken, Hendriks, & Van den Brink, 2002) was used to determine craving in three subscales: desire, negative
reinforcement, and (loss of) control.
To determine the levels of state and trait anxiety, the State- Trait Anxiety Inventory (STAI) (Spielberger, 1989) was used, and the Beck Depression Inventory (BDI) (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) was conducted to determine depression levels in participants. To determine cannabis use, the Cannabis Use Disorders Identification Test (CUDIT) was performed to check for any poly-drug use (Loflin, Babson, Browne, & Bonn-Miller, 2018).
Behavioral data analysis
To determine main and interaction effects of group (controls and cocaine users) and sex (men and women) mostly two-way ANOVA tests with
interaction effect were executed using R. If there appeared to be an interaction effect between sex and cocaine use, planned one-way ANOVAs were conducted to determine the nature of the interaction effect. If either the assumption of normality or the assumption of homogeneity of
variance was violated, firstly a log transformation was conducted, and if the assumption of normality and the assumption homogeneity of variance were still violated, instead of a two-way ANOVA, a generalized linear model was conducted. A generalized linear model analysis is not as dependent on the assumption of normality and the assumption of
homogeneity of variance, therefore making it a good alternative to a two-way ANOVA. The two-two-way robust ANOVA was not used in this research, since this statistical test did not give enough information to determine effect size. For a one-way ANOVA, a robust ANOVA was performed
(O’Brien, 2012). A robust ANOVA is less dependent of the assumptions of normality and homogeneity of variance, thus making it a good alternative when these assumptions are violated. To determine significant differences between frequencies of the groups, chi-square tests were performed.
Structural MRI data acquisition and pre-processing
Images were acquired on a 3.0-T Philips Achieva DS scanner (Philips Medical Systems, Best, the Netherlands) using a 32-channel head coil. A T1-3D anatomical scan (TR/TE 8.2/3.8; matrix 240 x 240; 1 x 1 mm3 voxel;
transverse slices) was taken.
To analyse the structural MRI scans, voxel-based morphometry (VBM) was performed using SPM12 (Penny W., Friston K., Ashburner, J., KIebel, S., Nichols, 2006) in MatLab R2016a with the toolbox CAT12 (Gaser, R., 2016). All of the participants’ images were reoriented using the anterior commissure as the origin. Pre-processing was done using mostly the toolbox CAT12 and consisted of segmentation (separation of
cerebrospinal fluid, grey and white matter), an estimation of total
intracranial volume (TIV), which would be used as a covariate in analysis along with age, and lastly, images were smoothed with an SPM12 tool, with a Gaussian kernel of 8 mm.
Statistical analyses
To determine the effects of CTQ on sex and group differences, a full factorial model was used in SPM12, where TIV, age, BDI and STAI were used, and a full factorial model was used where CTQ was used as a covariate. On top of that, to determine a three-way interaction effect of sex, group and CT, CTQ was used as a factor instead of as a covariate, with again TIV, age, BDI and STAI as covariates. WFU pickatlas was used to determine which brain regions were indicated with the cluster
coordinates (Maldijan, J.A., Laurienti, P.J., Burdette, 2004) and MarsBar was used to extract data of volumes of specific brain regions (Brett, M., Anton, J.L., Valabregue. R., Poline, 2002).
Results CU W(n=26) CU M(n=31) NC W(n=28) NC M(n=28) Main effect group Main effect sex Interaction effect Age 26.65 ± 6.82 29.19 ± 8.14 26.32 ± 5.00 25.93 ± 5.70 p = 0.101F1,109 = 2.74 ω2 = 0.015 p = 0.352 F1,109 = 0.87 ω2 = -0.001 p = 0.208 F1,109 = 1.61 ω2 = 0.005 STAI score* 76.50 ± 14.08 68.61 ± 18.09 66 ± 13.4 65.92 ± 13.38 p = 0.040 F1,109 = 4.33 ω2 = 0.028 p = 0.123 F1,109 = 2.39 ω2 = 0.012 p = 0.171 F1,109 = 1.90 ω2 = 0.008 BDI score* 14.58 ± 8.37 8.97 ± 8.57 4.50 ± 5.22 5.04 ± 4.73 p < 0.001 F1,89 = 23.14 ω2 = 0.175 p = 0.229 F1,89 = 1.47 ω2 = 0.004 p = 0.091 F1,89 = 2.92 ω2 = 0.015 CTQ score* 51.04 ± 10.171,2 48.10 ± 8.271 46.00 ± 7.812 44.78 ± 4.47 p = 0.022 F = 5.62 ξ = 0.33 p = 0.687 F = 0.16 ξ = 0.06 1 = p = 0.002 z = -3.55 η2 = 0.111 2 = p <0.001 z = 4.58 η2 = 0.185 Childhood trauma (CT) 46.15% (n=12)1 29.03% (n=9) 14.29% (n=4)1 14.29% (n=4) p = 0.011 χ2 = 6.40 η2 = 0.0968 p = 0.479 χ2 = 0.50 η2 = 0.009 1 = p = 0.017 χ2 = 8.14 η2 = 0.200 Cocaine use (g/m)* 3.59 ± 5.08 5.34 ± 5.52 - - - p = 0.122F1,50 = 2.58 ω2 = 0.009 -Cocaine onset (age) 19 ± 3.43 20 ± 5.10 - - - p = 0.122Q1,30.95 = 2.13 ξ = 0.290 -DDQ (total)* 34.58 ± 14.27 38.00 ± 12.68 - - - p = 0.087F1,55 = 3.11 ω2 = 0.003 -- Desire 14.23 ± 8.89 15.23 ± 8.89 - - - p = 0.326Q1,32.11 = 0.99 ξ = 0.170 -- Negative reinforc ement* 9.96 ± 6.24 12.06 ± 6.10 - - - p = 0.162F1,55 = 2.05 ω2 = 0.026 -- (Loss of) Control* 10.38 ± 3.72 10.71 ± 4.56 - - - p =0.485F1,55 = 0.50 ω2 = -0.018 -DUDIT score 16.34 ± 4.74 16.87 ± 4.86 - - - p = 0.716F1,55 = 0.134
-ω2 = -0.015 AUDIT score 12.35 ± 3.89 10.65 ± 5.79 4.54 ± 3.04 3.79 ± 3.17 p < 0.001 F = 85.61 ξ = 0.88 p = 0.246 F = 1.37 ξ = 0.16 ** CUDIT score 4.76 ± 7.68 6.22 ± 7.80 - - - p = 0.63F1,28.03= 0.63 ξ = 0.15
-Table 1. Demographic and clinical assessment information. Included values are p-value, test statistics, and effect size. If an interaction effect is found, it will be shown where the interaction effect lies with respectively a 1 or a 2.
*: The values that are shown here are the actual test scores, but for analysis a log transformation was used during analysis to account for outliers and assumptions of normality and homogeneity of variance.
**: All of the interactions between group by sex were not significant here, but since chi-square tests and generalized linear models do not give out one single interaction effect, the interaction effects individually can be found in Appendix 1.
Covariates
All of the information about the demographic and clinical assessment can be found in Table 1, including p-values, test statistics, and effect size. Two cocaine using women were excluded from the research because of missing MRI data. A two-way ANOVA demonstrated that there was no main or interaction effect of group and sex on age. For the State- and Trait Anxiety Inventory however, an ANOVA showed that STAI scores were significantly higher in CU compared to NC, but there were no main sex nor sex by group interaction effects. For BDI, an ANOVA showed that BDI scores were significantly higher in CU than in NC, but again no difference was found for main sex nor a sex by group interaction effects.
Childhood trauma
The CTQ scores were significantly higher in CU compared to NC using a generalized linear model. On top of that, CTQ scores were significantly higher in CU women compared to NC women, and CU women had
significantly higher scores than CU men, meaning CU women showed the highest amount of childhood trauma (CT) severity in all groups.
Additionally, a higher percentage of severe CT categorization was found in CU compared to NC. Also, there were significantly more CU women than NC women that were categorized with severe CT, but no main sex nor other interaction effects were found when using a chi-square test.
Exploratory variables
The AUDIT scores demonstrated that CU significantly scores higher than NC, but no main sex nor interaction effect was found when using a
generalized linear model.
The following results describe sex differences within cocaine users on cocaine-use variables.
Cocaine use per month was not significantly different between men and women. 1 participant was excluded from this questionnaire as they were
an outlier. Age of onset of cocaine use also did not significantly differ between sexes. 4 participants were excluded from age of onset analysis as they did not complete the questionnaire. On top of that, one-way
ANOVAs demonstrated that the DDQ score also did not significantly differ between men and women, and neither did the sub questionnaires (desire, negative reinforcement and (loss of) control). The DUDIT and the CUDIT also did not exhibit significantly different scores between men and
women.
Structural MRI analysis
Comparing group and sex
To determine a group and sex interaction without a correction for CT, CTQ was not used as a covariate. The results of this analysis can be found in Table 2. Cluste r size (Numb er of voxels) MNI coordinates (peak) Z value
(voxels) Brain region Cluster p value
Main effect group - CU > NC No signific ant cluster s - NC > CU 686 -57 -50 -34 4.31 Cerebellum P < 0.001 Main effect sex
- M > W 390 5274 689 8 26 -26 0 -94 -10 -4 -42 9 4.33 4.63 4.17 Right Caudate Nucleus Cerebellum Left putamen P = 0.011 P < 0.001 P < 0.001 - W > M 617 16 -2 8 4 33 56 4.37 3.71 Right middle cingulate cortex Left supplementary motor area P < 0.001 Sex x Group interaction - Men (NC > CU) > Women (NC > CU) 344 -2 2 56 3.92 Left supplementary motor area P = 0.003 - Women (NC > CU) > Men (NC > CU) No signific ant cluster s
Table 2. Differing brain regions between men, women, cocaine users and non-drug using controls. Shown here respectively are: cluster size of the significant brain region, coordinates of the peak (MNI)
of the brain region, the Z value of the voxels, the concerned brain region and the cluster p value.
There were no brain regions that were smaller for NC compared to CU. However, CU did show a smaller cerebellum than NC. Also, women did show a smaller right caudate nucleus, cerebellum and left putamen compared to men, while men showed a smaller right middle cingulate cortex (right MCC) and left supplementary motor area (left SMA)
compared to women. On top of that, a sex x group interaction was found in the left SMA.
Comparing group and sex without childhood trauma
To only compare a group and sex interaction, CTQ was used as a covariate to correct for CT effects on the brain. The results of this analysis can be found in Table 3 and Figure 1-5.
Cluster size (Number of voxels) MNI coordinates (peak) Z value (voxel s)
Brain region Cluster p value Main effect group - CU > NC No significant clusters - NC > CU 555 -57 -50 -34 4.17 Cerebellum P = 0.001 Main effect sex
- M > W 354 2032 716 8 -21 -26 0 -56 -10 -4 -36 9 4.32 4.38 4.13 Right Caudate Nucleus Cerebellum Left Putamen P = 0.018 P < 0.001 P < 0.001 - W > M 701 16 8 36 4.33 Left supplementary motor area P < 0.001 Sex x Group interaction - Men (NC > CU) > Women (NC > CU) 507 -2 2 56 3.94 Left supplementary motor area P = 0.002 - Women (NC > CU) > Men (NC No significant clusters
> CU)
Table 3. Differing brain regions between men, women, cocaine users and non-drug using controls. Shown here are respectively: cluster size of the significant brain region, coordinates of the peak (in MNI) of the brain region, the Z value from the voxels, the concerned brain region and the cluster p value.
Figure 1-5. Brain regions that differ between cocaine users, non-drug using controls, men and women. A: Cocaine users show a
smaller cerebellum compared to non-drug using controls. B/C/D: Women show a smaller right caudate nucleus, cerebellum and putamen than men. E: Men show a smaller left SMA than women. F: CU women show a bigger left SMA than NC women, CU M show a bigger left SMA than NC men, and CU women show a bigger left SMA than CU men.
There were no brain regions that were significantly smaller for the NC group compared to the CU group. However, the cerebellum was
significantly smaller for the CU group compared to the NC group. On top of that, the right caudate nucleus, the left putamen) and the cerebellum were smaller for the women group compared to the men group. Women did show a bigger left supplementary motor area (left SMA) compared to men. When looking at an interaction effect between group and sex, there was a significant difference between group and sex, which is investigated more in-depth in Table 4.
For left SMA interaction effects, 3 interaction effects were found. Firstly, CU men had significantly smaller left SMA volumes compared to CU women (cluster size = 338, p < 0.001). Secondly, NC women had
significantly smaller left SMA volumes compared to CU women (cluster size = 15, p = 0.026). Lastly, NC men had significantly smaller left SMA volumes compared to CU men (cluster size = 6, p = 0.047). No other interaction effects were found. However, when comparing cluster sizes between all the significant interaction effects, CU women > CU men did show a way higher cluster size and was way more significant than the other two comparisons, making this the most important result, making up much of the interaction effect in the left SMA.
Graph 1. Differences in left SMA volume in CU, NC, men and
women. Shown here are the mean brain volumes with standard deviation error bars of the left supplementary cortex. While not a lot of effect is shown in the graph itself, a p value of < 0.001(**) has been observed between CU women and CU men, where CU men show significantly lower volumes in the left supplementary motor cortex than CU women.
When comparing the analyses with and without CT as a correction, most brain regions appear in both analyses, implying that CT does not
correlate with these regions (cerebellum, right caudate nucleus, left putamen, and left SMA). However, the right MCC disappears in analysis when correcting for CTQ, giving an indication that CT somehow
correlates with the right MCC, and more importantly, men demonstrate a negative correlation with CT in the right MCC, while women show a positive correlation with CT in the right MCC.
Graph 2. Differences in right MCC volume in CU, NC, men and women. Shown here are the mean brain volumes with standard deviation error bars of the right middle cingulate cortex. A main sex effect of
p<0.001 (**) has been found.
Comparing group, sex and childhood trauma
To determine group by sex by CT interaction, CTQ was used as a factor in analysis. The results of this analysis can be found in Table 5 and Figure 6-9. Cluster size (number of voxels) MNI coordinates (peak) Z value
(voxels) Brain region Clusterp value CTQ - Positive correlat ion 386 1439 4 -45 -22 18 32 40 3.48 3.77 Right middle cingulate cortex Left middle frontal cortex P = 0.013 P < 0.001 - Negativ e correlat ion 513 -38 8 -32 4.11 Left superior/middle temporal pole P = 0.002 CTQ x Group No significant clusters CTQ x Sex - Women > Men 406
-30 22 62 4.23 Left middle frontal lobe P = 0.02
- Men > Women No significant clusters CTQ x Group x Sex No significant clusters
Table 5. Differing brain regions between men women, cocaine users and non-drug using controls and childhood trauma as an interaction covariate. Shown here are respectively: cluster size of the significant brain region, coordinates of the peak (in MNI) of the brain region, the Z value from the voxels, the concerned brain region, and the cluster p-value
It was found that CT positively correlates with the right MCC (although this effect is explained through a CT by sex effect) and the left middle frontal cortex (left MFC), it correlates negatively with the left
superior/middle temporal pole, showing the effects of only CT on the brain. No CT by group effects were found to be significant, but CT by sex showed the left middle frontal lobe to be bigger in women than in men. No three-way interactions of CT, group and sex were found to be
significant.
Figure 6-9. Brain regions that have an interaction effect with childhood trauma, comparing childhood trauma, cocaine users and non-drug using controls, women and men. A: Childhood trauma shows a positive correlation to the grey matter of the right middle cingulate cortex. B: Childhood trauma shows a positive correlation with the left middle frontal cortex. C: Childhood trauma shows a negative correlation with the left
superior/middle temporal pole. D: Men with an interaction effect with childhood trauma show a smaller left middle frontal lobe.
Graph 3. Differences in left MFC volume in CU, NC, men and
women. Shown here are the mean brain volumes with standard deviation error bars of the left middle frontal cortex. A main sex effect of of p < 0.05 (*) has been found.
Discussion
The aim of this study was to investigate whether sex differences in
cocaine use could be partially influenced by childhood trauma (CT). When looking at sex differences only, using a correction of CT, it was
demonstrated that women showed a smaller grey matter volume in the right caudate nucleus, cerebellum and left putamen, men showed a
smaller volume in the left SMA than women. On top of that, an interaction effect was found, showing that cocaine using (CU) men showed a smaller left SMA than CU women. While correcting for CT, it was found that the right middle cingulate cortex (MCC) and the left middle frontal cortex (MFC) had a negative correlation with CT in men, and a positive
correlation with CT in women. Finally, no CT by group interaction was found, and no CT by sex by group interaction was found either.
What can be concluded from this research is that CT does not partially explain sex differences in grey matter morphometry in cocaine users. However, the right MCC and the left MFC do show different correlations with grey matter volume in men and women, only this was not cocaine-dependent, even though this was not expected.
On top of that, it has been shown that there are sex differences in brain region volumes which were not cocaine- or CT-dependent. Moreover, CT does negatively correlate with the left superior/middle temporal lobe, which was expected. None of the before mentioned results seemed to be explainable by the demographic data since only CTQ scores showed an interaction effect between cocaine-using women and men, but showed higher scores for women. The exploratory research that determines
severity of addiction to cocaine, alcohol and cannabis also did not explain the unexpected results.
Most of these afore mentioned results go against the hypothesis of this research, which states that sex differences in cocaine users could be explained partially through CT, where women would show bigger brain region abnormalities in grey matter morphometry compared to men in affected brain regions by CT and cocaine use. Men would still be affected by CT compared to controls, but not in the same degree as women would. Even though the caudate nucleus and the putamen showed a smaller grey matter volume in men compared to women, this was not dependent of CT and was found as a main effect of sex. This goes against research of
Cohen et al. (2006), whom showed a decrease in grey matter volume in the caudate nucleus in rodents with CT. It also goes against the
demographical results in this study, which states that CT prevalence is higher in women compared to men. However, research of Abedelahi, Hasanzadeh, Hadizadeh, & Joghataie (2013) showed that women did show a bigger caudate nucleus and putamen than men, where it was stated that these morphometric changes could be caused due to hormonal differences between sexes.
Also, CU women showed a larger volume in the left SMA compared to CU men, which was not expected. Even though, research of Pletzer et al. (2010) also shows a bigger left supplementary cortex on non-drug using women than men. Even though they did not look at cocaine users, in this
research no difference was found between CU and non-drug using controls (NC).
An interaction effect between group and sex was also found in the left SMA. The volume of the lest SMA was larger in CU women compared to CU men. This interaction effect most probably explains the interactions effect, leaving the other two interaction effects that were found out of the discussion.
Even though this effect was not expected, research of Pletzer et al. (2010) also shows a bigger left supplementary cortex on non-drug using women than men. Even though they did not look at cocaine users, no difference between CU and NC was found in this study. In this paper they also look at the hormonal cycle and women who take contraceptives with estrogen and progesterone. The women who take contraceptives show a bigger left SMA and other regions compared to women without contraceptives, and women without contraceptives also show a bigger left SMA than men. This shows that hormones could play a big role in the differences in development between men and women (Pletzer et al., 2010; Witte, Savli, Holik, Kasper, & Lanzenberger, 2010) and possibly explain sex
differences in cocaine use (Jackson, Robinson, & Becker, 2006), causing differences in grey matter morphometry during development and could possibly influence the initiation of addiction and how an individual reacts to an addiction.
The results showed that the left MFC and the right MCC were both
positively correlated with CT in women and negatively correlated with CT in men. This went again previous research, where it was stated that generally CT is linked with negative correlations with brain regions (Cohen et al., 2006; Farrell et al., 2016; Gorka et al., 2014; Jackowski et al., 2011; Loi et al., 2014), most notably in the hippocampus, orbitofrontal cortex, insula, prefrontal cortex and the caudate nucleus. What could possibly explain these differences are two explanations:
Firstly, women showing a positive correlation in brain volumes with CT could be caused by sex hormones as mentioned before: it had been shown that more oestrogen and progesterone are correlated with larger volumes in brain regions (Pletzer et al., 2010). During puberty, oestrogen and progesterone levels are different than normal. Because CT development takes place before puberty, this could result in a positive correlation in brain regions like the left MFC and the right MCC. Secondly, an
explanation why women are positively correlated and men are negatively correlated with CT in the right MCC and the left MFC is that there is also a possibility that women are more neurobiologically resilient compared to men (De Bellis & Keshavan, 2003; Samplin, Ikuta, Malhotra, Szeszko, & DeRosse, 2013). Previous research did show a higher neurobiological resilience in women with PTSD compared to men with PTSD. What this could mean, is that in men neurobiological changes are more noticeable compared to women, since they go through less brain volume changes when confronted with, in this case, CT. This could also possibly explain the unexpected disconnect between brain volume results and
demographical results concerning CT, as it is also shown in previous research that women do tend to have a higher prevalence in CT (Stein,
Walker, & Forde, 2000). This exhibits that sex differences could be partially explained by other factors, including sex hormones and neurobiological resilience.
The negative correlation of CT with the left superior/middle temporal pole did confirm expectations, and these results were also found in Gold et al. (2016). As the superior/middle temporal pole is involved in emotion
processing, a smaller left superior/middle temporal pole would be
associated with abnormalities in emotional processing, which is the case in CT.
To conclude, it was shown that CT correlated with sex differences, even though no effect was found for cocaine use. This could possibly be explained through sex hormones and neurobiological resilience. Sex hormones were included in the overall project, but were unfortunately not used in this study, so the role of sex hormones in CT correlation with sex cannot be confirmed.
One of the reasons why there were no differences between CU and NC could be the fact that in this research, regular cocaine users were recruited which did not all fit a DSM-V diagnosis (American Psychiatric Association, 2013). This could have diminished effects that would
normally be there between CU and NC. Another thing that stood out was that CU filled in a way higher amount of cocaine during the initial online screening compared to when they were asked again how much cocaine they used when they were at the test centre. This could be caused by the fact that they became aware of their usage of cocaine and used less that month, or they did not complete the questionnaires realistically. Initially the timeline follow-back questionnaire was used. However, a lot of CU participants did not complete this questionnaire, making it unable for use. Another cocaine-use questionnaire was used, although this list was a little less accurate compared to the previous one.
What could also have affected the results of this research was the fact that most of the CU participants in our research were poly-drug users, making it hard to determine if effects were caused by cocaine use or that they were caused because of poly-drug use. But since correcting for even more covariates would reduce statistical power, this was not thoroughly investigated.
One of the limitations of this research was a selection bias: most of the participants were from or around Amsterdam, making it a very selective group of people, therefore making it not a random sample of the
population. Amsterdam is a city where drugs are used more often than in other cities in the Netherlands (Europees Drugs rapport, 2019), so it does not give a general idea of cocaine use.
One of the strengths of this research was that compared to other cocaine dependence research, this research had relatively a lot of participants. A lot of other research does not have a very big cocaine-using group, and this research has almost 60 cocaine-using participants (Adinoff et al., 2006; Ernst, Chang, Oropilla, Gustavson, & Speck, 2000; Franklin et al., 2002).
Another strength was that prior to this research, not a lot of research had been done about sex differences in substance and cocaine use and the role of CT in developing of a cocaine use dependence. Even though this made it hard to have strong expectations before researching, this
research shows it is important to not leave out women completely when looking at cocaine addiction.
This research shows it is important to investigate sex differences in cocaine use. In this study, it has been shown that there are clear
differences in grey matter morphometry between men and women. This confirms what psychological research stated before: women and men tend to react differently to addiction, which could partially be explained through these sex differences in grey matter volume. This study shows the importance of investigating sex differences in substance use. More knowledge about sex differences should lead to more efficient
individualized treatment programs for men and women.
For follow-up research, it would be interesting to look at the connectivity of the fronto-striatal network system for further research, as in this
research, a lot of brain regions falling under the fronto-striatal network are affected. Looking at the connectivity would maybe explain some of the unexpected results that were found when looking at cocaine use and sex differences. It would also be interesting to investigate fronto-striatal network and hormonal changes, as that seems to have a lot of impact on differing development of the brain for men and women.
All in all, this research has shown that there are differences between men and women, and women should be taken into consideration more when it comes to research and individual treatment plans.
References
Abedelahi, A., Hasanzadeh, H., Hadizadeh, H., & Joghataie, M. T. (2013). Morphometric and volumetric study of caudate and putamen nuclei in normal individuals by MRI: Effect of normal aging, gender and
hemispheric differences. Polish Journal of Radiology, 78(3), 7–14. https://doi.org/10.12659/PJR.889364
Adinoff, B., Williams, M. J., Best, S. E., Harris, T. S., Chandler, P., & Devous, M. D. (2006). Sex differences in medial and lateral
orbitofrontal cortex hypoperfusion in cocaine-dependent men and women. Gender Medicine, 3(3), 206–222.
https://doi.org/10.1016/S1550-8579(06)80209-3
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). In Washington, DC.
Anker, J.J., Carroll, M. E. (2010). Females Are More Vulnerable to Drug Abuse than Males Evidence from Preclinical Studies and the Role of Ovarian Hormones. Behavioral Neurosciences, 8.
BECK, A. T., WARD, C. H., MENDELSON, M., MOCK, J., & ERBAUGH, J. (1961). An Inventory for Measuring Depression. Archives of General
Psychiatry, 4(6), 561–571.
https://doi.org/10.1001/archpsyc.1961.01710120031004
Beck, A., Wüstenberg, T., Genauck, A., Wrase, J., Schlagenhauf, F., Smolka, M. N., … Heinz, A. (2012). Effect of Brain
Structure..BrainFunction,andBrain Connectivity on Relapse in Alcohol-Dependent Patients. Archives of General Psychiatry, 69(8), 842–853. https://doi.org/10.1001/archgenpsychiatry.2011.2026 Becker, J. B., & Koob, G. F. (2016). Sex differences in animal models:
Focus on addiction. Pharmacological Reviews, 68(2), 242–263. https://doi.org/10.1124/pr.115.011163
Berman, A. H., Bergman, H., Palmstierna, T., & Schlyter, F. (2003). The
Drug Use Disorders Identification Test: Manual. 1–16.
Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D.,
Ahluvalia, T., … Zule, W. (2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child
Abuse and Neglect, 27(2), 169–190.
https://doi.org/10.1016/S0145-2134(02)00541-0
Bobzean, S. A. M., DeNobrega, A. K., & Perrotti, L. I. (2014). Sex differences in the neurobiology of drug addiction. Experimental
Neurology, 259, 64–74.
https://doi.org/10.1016/j.expneurol.2014.01.022
Brett, M., Anton, J.L., Valabregue. R., Poline, J. B. (2002). Region of interest analysis using an SPM toolbox. NeuroImage, 16(2).
Casey, B. J., & Jones, R. M. (2010). Neurobiology of the adolescent brain and behavior: Implications for substance use disorders. Journal of the
American Academy of Child and Adolescent Psychiatry, 49(12), 1189–
1201. https://doi.org/10.1016/j.jaac.2010.08.017
Cohen, R. A., Grieve, S., Hoth, K. F., Paul, R. H., Sweet, L., Tate, D., … Williams, L. M. (2006). Early Life Stress and Morphometry of the Adult Anterior Cingulate Cortex and Caudate Nuclei. Biological
https://doi.org/10.1016/j.biopsych.2005.12.016
De Bellis, M. D., & Keshavan, M. S. (2003). Sex differences in brain maturation in maltreatment-related pediatric posttraumatic stress disorder. Neuroscience and Biobehavioral Reviews, 27(1–2), 103–117. https://doi.org/10.1016/S0149-7634(03)00013-7
Elliott, D. M., Mok, D. S., & Briere, J. (2004). Adult sexual assault: Prevalence, symptomatology, and sex differences in the general population. Journal of Traumatic Stress, 17(3), 203–211.
https://doi.org/10.1023/B:JOTS.0000029263.11104.23
Enoch, M. A., Hodgkinson, C. A., Yuan, Q., Shen, P. H., Goldman, D., & Roy, A. (2010). The Influence of GABRA2, Childhood Trauma, and Their Interaction on Alcohol, Heroin, and Cocaine Dependence.
Biological Psychiatry, 67(1), 20–27.
https://doi.org/10.1016/j.biopsych.2009.08.019
Ernst, T., Chang, L., Oropilla, G., Gustavson, A., & Speck, O. (2000). Cerebral perfusion abnormalities in abstinent cocaine abusers: A perfusion MRI and SPECT study. Psychiatry Research -
Neuroimaging, 99(2), 63–74.
https://doi.org/10.1016/S0925-4927(00)00056-1
Europees Drugs rapport. (2019).
Farrell, M. R., Holland, F. H., Shansky, R. M., & Brenhouse, H. C. (2016). Sex-specific effects of early life stress on social interaction and
prefrontal cortex dendritic morphology in young rats. Behavioural
Brain Research, 310, 119–125.
https://doi.org/10.1016/j.bbr.2016.05.009
Franken, I. H. A., Hendriks, V. M., & Van den Brink, W. (2002). Initial validation of two opiate craving questionnaires: The Obsessive
Compulsive Drug Use Scale and the Desires for Drug Questionnaire.
Addictive Behaviors, 27(5), 675–685.
https://doi.org/10.1016/S0306-4603(01)00201-5
Franklin, T. R., Acton, P. D., Maldjian, J. A., Gray, J. D., Croft, J. R., Dackis, C. A., … Childress, A. R. (2002). Decreased Gray Matter
Concentration in the Insular , Cocaine Patients. Biological Psychiatry,
51(01), 134–142.
Gaser, R., D. (2016). A Computational Anatomy Toolbox for the Analysis of Structural MRI Data. HBM 2016.
Gold, A. L., Sheridan, M. A., Peverill, M., Busso, D. S., Lambert, H. K., Alves, S., … McLaughlin, K. A. (2016). Childhood abuse and reduced cortical thickness in brain regions involved in emotional processing.
Journal of Child Psychology and Psychiatry and Allied Disciplines, 57(10), 1154–1164. https://doi.org/10.1111/jcpp.12630
Gorka, A. X., Hanson, J. L., Radtke, S. R., & Hariri, A. R. (2014). Reduced hippocampal and medial prefrontal gray matter mediate the
association between reported childhood maltreatment and trait anxiety in adulthood and predict sensitivity to future life stress.
Biology of Mood & Anxiety Disorders, 4(1), 12.
https://doi.org/10.1186/2045-5380-4-12
Holmes, A. J., Hollinshead, M. O., Roffman, J. L., Smoller, J. W., & Buckner, R. L. (2016). Individual differences in cognitive control circuit
Journal of Neuroscience, 36(14), 4038–4049.
https://doi.org/10.1523/JNEUROSCI.3206-15.2016
Hyman, S. M., Paliwal, P., Chaplin, T. M., Mazure, C. M., J, B., & Sinha, R. (2009). Outcomes in Women but not Men. Alcohol, 92, 208–216.
Jackowski, A., Perera, T. D., Abdallah, C. G., Garrido, G., Tang, C. Y., Martinez, J., … Kaufman, J. (2011). Early-life stress, corpus callosum development, hippocampal volumetrics, and anxious behavior in male nonhuman primates. Psychiatry Research - Neuroimaging, 192(1), 37– 44. https://doi.org/10.1016/j.pscychresns.2010.11.006
Jackson, L. R., Robinson, T. E., & Becker, J. B. (2006). Sex differences and hormonal influences on acquisition of cocaine self-administration in rats. Neuropsychopharmacology, 31(1), 129–138.
https://doi.org/10.1038/sj.npp.1300778
Koob, G. F. (2009). Dynamics of neuronal circuits in addiction: reward, antireward, and emotional memory. Pharmacopsychiatry, 42 Suppl 1, 32–41. https://doi.org/10.1055/s-0029-1216356
Kosten, T. A., Sanchez, H., Zhang, X. Y., & Kehoe, P. (2004). Neonatal isolation enhances acquisition of cocaine self-administration and food responding in female rats. Behavioural Brain Research, 151(1–2), 137–149. https://doi.org/10.1016/j.bbr.2003.08.010
Lawford, B. R., Barnes, M., Connor, J. P., Heslop, K., Nyst, P., & Young, R. M. D. (2012). Alcohol use disorders identification test (AUDIT) scores are elevated in antipsychotic-induced hyperprolactinaemia. Journal of
Psychopharmacology, 26(2), 324–329.
https://doi.org/10.1177/0269881110393051
Loflin, M., Babson, K., Browne, K., & Bonn-Miller, M. (2018). Assessment of the validity of the CUDIT-R in a subpopulation of cannabis users.
American Journal of Drug and Alcohol Abuse, 44(1), 19–23.
https://doi.org/10.1080/00952990.2017.1376677
Loi, M., Koricka, S., Lucassen, P. J., & Joëls, M. (2014). Age- and sex-dependent effects of early life stress on hippocampal neurogenesis.
Frontiers in Endocrinology, 5(FEB), 1–11.
https://doi.org/10.3389/fendo.2014.00013
Lynch, W. J., Roth, M. E., & Carroll, M. E. (2002). Biological basis of sex differences in drug abuse: Preclinical and clinical studies.
Psychopharmacology, 164(2), 121–137.
https://doi.org/10.1007/s00213-002-1183-2
Makris, N., Gasic, G. P., Seidman, L. J., Goldstein, J. M., Gastfriend, D. R., Elman, I., … Breiter, H. C. (2004). Makris et al 2004. 44, 1–12.
Retrieved from
papers2://publication/doi/10.1016/j.neuron.2004.10.027
Maldijan, J.A., Laurienti, P.J., Burdette, J. H. (2004). Precentral gyrus discrepancy in electronic versions of the Talairach atlas. NeuroImage
2004, 21, 450–455.
O’Brien, R. G. (2012). A General ANOVA Method for Robust Tests of Additive Models for Variances. Journal of the American Statistical
Association, 74(368), 877–880.
Penny W., Friston K., Ashburner, J., KIebel, S., Nichols, T. (2006). Statistical Parametric Mapping: The Analysis of Functional Brain Images (1st ed.). In Academic press.
Pletzer, B., Kronbichler, M., Aichhorn, M., Bergmann, J., Ladurner, G., & Kerschbaum, H. H. (2010). Menstrual cycle and hormonal
contraceptive use modulate human brain structure. Brain Research,
1348, 55–62. https://doi.org/10.1016/j.brainres.2010.06.019
Rhemtulla, M., Fried, E. I., Aggen, S. H., Tuerlinckx, F., Kendler, K. S., & Borsboom, D. (2016). Network analysis of substance abuse and
dependence symptoms. Drug and Alcohol Dependence, 161, 230–237. https://doi.org/10.1016/j.drugalcdep.2016.02.005
Saleh, A., Potter, G. G., McQuoid, D. R., Boyd, B., Turner, R., MacFall, J. R., & Taylor, W. D. (2017). Effects of early life stress on depression, cognitive performance and brain morphology. Psychological Medicine,
47(1), 171–181. https://doi.org/10.1017/S0033291716002403
Samplin, E., Ikuta, T., Malhotra, A. K., Szeszko, P. R., & DeRosse, P. (2013). Sex differences in resilience to childhood maltreatment: Effects of trauma history on hippocampal volume, general cognition and subclinical psychosis in healthy adults. Journal of Psychiatric
Research, 47(9), 1174–1179.
https://doi.org/10.1016/j.jpsychires.2013.05.008
Spielberger, C. D. (1989). State- Trait Anxiety Inventory: Bibliography (2nd ed.). In Palo Alto, CA: Consulting Psychologists Press.
Stein, M. B., Walker, J. R., & Forde, D. R. (2000). Gender differences in susceptibility to posttraumatic stress disorder. Behaviour Research
and Therapy, 38(6), 619–628.
https://doi.org/10.1016/S0005-7967(99)00098-4
Teicher, M. H., Anderson, C. M., & Polcari, A. (2012). Childhood
maltreatment is associated with reduced volume in the hippocampal subfields CA3, dentate gyrus,and subiculum. Proceedings of the
National Academy of Sciences of the United States of America, 109(9). https://doi.org/10.1073/pnas.1115396109
van Dam, D., Ehring, T., Vedel, E., & Emmelkamp, P. M. G. (2010). Validation of the Primary Care Posttraumatic Stress Disorder
screening questionnaire (PC-PTSD) in civilian substance use disorder patients. Journal of Substance Abuse Treatment, 39(2), 105–113. https://doi.org/10.1016/j.jsat.2010.05.005
Witte, A. V., Savli, M., Holik, A., Kasper, S., & Lanzenberger, R. (2010). Regional sex differences in grey matter volume are associated with sex hormones in the young adult human brain. NeuroImage, 49(2), 1205–1212. https://doi.org/10.1016/j.neuroimage.2009.09.046 Yang, G., Sau, C., Lai, W., Cichon, J., & Li, W. (2015). Emergence of Sex
Differences in the Development of Substance Use and Abuse during Adolescence. 344(6188), 1173–1178.
Appendix 1. interaction effects in childhood trauma subcategories CU W – CU M HC W – HCM CU W – HCW CU M – HCM Emotional abuse p = 0.376 χ2 = 15.04 η2 = 0.261 p = 0.300 χ2 = 11.78 η2 = 0.261 p = 0.331§ χ2 = 15.72 η2 = 0.305 p = 0.042 χ2 = 20.253 η2 = 0.338 Emotional neglect p = 0.320χ2 = 0.99 η2 = 0.033 p = 0.967 χ2 = 0.01 η2 = 0.001 p = 0.136 χ2 = 2.22 η2 = 0.073 p = 0.508 χ2 = 0.44 η2 = 0.015 Physical abuse p = 0.157χ2 = 2.01 η2 = 0.064 p = 1 χ2 = 0 η2 = 0 p = 0.084 χ2 = 2.98 η2 = 0.095 p = 0.084 χ2 = 2.98 η2 = 0.095 Physical neglect p = 0.290χ2 = 1.12 η2 = 0.037 p = 0.604 χ2 = 0.27 η2 = 0.010 p = 0.163 χ2 = 1.95 η2 = 0.065 p = 0.546 χ2 = 0.36 η2 = 0.001 Sexual abuse p = 0.157χ2 = 2.01 η2 = 0.064 p = 0.084 χ2 = 2.98 η2 = 0.092 p = 0.206 χ2 = 1.60 η2 = 0.054 p = 0.084 χ2 = 2.98 η2 = 0.088 Table 6. Interaction effects in childhood trauma subcategories.
Shown here are the p value, the test statistics and the effect size, respectively. CU W – CU M HC W – HC M CU W – HC W CU M – HC M AUDIT scores p = 0.231 z = -1.89 η2 = 0.1 p = 0.516 z = -1.37 η2 = -0.133 p < 0.001 z = 9.55 η2 = 0.93 p < 0.001 z = 9.26 η2 = 0.90
Table 7. Interaction effects in AUDIT scores. Included are p values, z statistics and effect sizes. The significance in CU W – HC W and CU M – HC M together make for a main group effect.
