Abstract
Poly-drug consumption is a dangerous, yet complex model of substance use that contributes to many cases of imprisonment and fatal overdose. Despite the growing number of studies looking at this phenomenon, there remains a lack of neuroimaging data elucidating the neural markers of poly-drug use. In particular, the valuation system, tasked with weighing the value of items and experiences, may hold significance in understanding the motivations behind poly-drug use. To this end, we sought to analyze the functional connectivity of the Valuation System (VS), Executive Control System (ECS) and Valuation-Control Complex (VCC) of drug-using participants in the Human Connectome Project Healthy Young Adult dataset (n=992). Using multivariate regression, network-averaged connectivities were correlated with various substance use measures (stimulants, cocaine, hallucinogens, opiates, sedatives and marijuana) and demographic variables (gender, parental use history). We found that the VS, ECS and VCC were all correlated with drug use behaviour either as individual systems or when paired with other substances. Both VS (R2= 0.53) and ECS (R2= 0.55) connectivity are positively correlated with stimulant use whereas both ECS (R2= 0.45) and VCC (R2= 0.045) connectivity are negatively correlated with marijuana use. VS (R2= 0.75) and VCC (R2= 0.74) connectivities are also negatively correlated with sedative use. Additionally, network connectivity was correlated with drug use behaviour via two-way interactions with other substances. These findings provide preliminary indications of the consequences of poly-drug use in healthy young adults.
Introduction
Poly-drug use is a growing problem worldwide with many rehabilitating patients having consumed more than one drug (Su et al., 2019; Wang et al., 2018). Poly-drug dependents mix-and-match different combinations in an attempt to increase the efficacy (Scott et al., 2007) or attenuate drawbacks (Fisk et al., 2011; Hernández-López et al., 2002) of a particular substance. Poly-drug use is related to dysfunctional behavioural regulation and executive functioning (Fernandez-Serrano et al., 2010; Keyes et al., 2008; Koester et al., 2013), as well as risky sexual practices (Floyd et al., 2010) and even imprisonment (Cooper et al., 2018). Despite the multi-layered nature of poly-drug use (Jongenelis et al., 2019), one concept could play a key role by underpinning a variety of substance use profiles: valuation.
Value is the currency of choice (Helie et al., 2017) and has been linked with drug use as seen in studies featuring heroin (Xie et al., 2014), stimulants (Gipson and Bardo, 2009), cocaine (Meade et al., 2011) and marijuana (Elton et al., 2019) dependents. Drugs become valued as a means of achieving a desired state, be it euphoria or relief from pain (Koob et al., 2014). As such, addictive drugs are sometimes considered valuable for their euphoria-inducing (Klega and Keehbauch, 2018) or pain-killing properties (Choi et al., 2017). Additionally aberrant executive control (Volkow et al., 2011; Xie et al., 2014; Zhai et al., 2015) has been observed during dependency. Research indicates that the relationship between valuation and executive control become unbalanced, possibly contributing to risky pursuit and consumption of drugs (Xie et al., 2014; Zhai et al., 2015). The valuation system (VS) is thought to be involved in estimating the incentive value of different options and comprises the ventromedial prefontal cortex, posterior cingulate cortex, amygdala and nucleus accumbens (van den Bos and McClure, 2013). The VS is implicated in value-setting (Ousdal et al., 2014) and is hypothesized to interface between pleasure and pain (Kringelbach, 2005; Rolls et al., 2003; Schultz, 2007; Seymour et al., 2005; Small et al., 2001; Tremblay and Schultz, 1999). The executive control system (ECS), consisting of the lateral prefrontal cortex, dorsal anterior cingulate cortex and posterior parietal cortex, guides behaviour in pursuit of long-term goals (McClure et al., 2004). The ECS may assist in refining VS-encoded values by incorporating elements of future-thinking and careful deliberation (McClure et al., 2004; van den Bos and McClure, 2013). Both the VS and ECS are linked with substance use behaviour both as a network (Xie et al., 2014; Zhai et al., 2015) and as individual regions (Bedi et al., 2015; Garavan et al., 2008; Kosten et al., 2006; Liu et al., 2009).
Although the concept of value has gained importance in the field of drug dependency research, functional neuroimaging evidence remains scarce. Relatively few studies have taken into account the wide range of drug types consumed concurrently (Jongenelis et al., 2019). Taken together with the tendency to exclude poly-drug users from individual substance use neuroimaging studies (Ma et al., 2011; Zhai et al., 2015), these limitations have lead a dearth in imaging-based findings. We aim to address this issue by correlating the connectivity of the VS, ECS and the valuation-control complex (VCC, a hybrid network comprised of VS and ECS regions) with poly-drug use in the Human Connectome Project Healthy Young Adult dataset using multivariate generalized linear models. We hypothesize that functional connectivity within the VS, ECS and VCC would all be correlated with alcohol, tobacco, opiates, cocaine, stimulants, sedatives and marijuana use. We further hypothesize that in some cases, connectivity may be correlated with substance use via a paired interaction with another drug (eg VS connectivity paired with sedative use may be negatively correlated with cocaine use).
Experimental Parameters
Participants, image acquisition and preprocessing
Minimally preprocessed fMRI data from the Human Connectome Project (HCP, Smith et al., 2013) was sourced for healthy adults of both genders (age range = 22-37). Recruitment procedures and inclusion/exclusion criteria are described elsewhere (Van Essen et al., 2013). Only subjects with all four repeated resting-state fMRI sessions (first and second scan sessions with left-right and right-left phase encoding directions), who also possessed complete delay discounting measures, were included (n = 992). The resting-state functional MRI (rsfMRI) data was acquired using the MGH-UCLA Connectome scanner (Smith et al., 2013). Briefly, 1200 consecutive volumes were acquired in ∼15 minutes in each scan sessions when the participants were in resting-state with eyes open. An optimized multi-band accelerated gradient-echo planar imaging protocol was adopted for high spatial and temporal resolution. Each subject was scanned twice over 2 days, and each scan consists of two rsfMRI sessions with the phase encoding direction of right-to-left and left-to-right The detailed parameters are as follow: Repetition time= 720 ms, Echo time=33.1 ms, Flip Angle= 52°, Field of view= 280 × 180 mm, Matrix of 140 × 90, Slice number=72 and the voxel size is 2 mm isotropic. The minimal preprocessing pipeline from the HCP team (Glasser et al., 2013; Smith et al., 2013) was adopted in this study.
VS, ECS and VCC mask preparation
The VS, ECS and VCC were delineated using the binary masks that combined regions of interest (ROIs) from both the Desikan-Killiany (Desikan et al., 2006) and Destrieux (Destrieux et al., 2010) parcellations. While most of the regions comprising the VS, ECS and VCC were assembled using the Desikan-Killiany atlas, the more refined parcellation of Destrieux atlas was useful in isolating the dACC. Specifically, for the VS, the total vmPFC was composed of two regions, the sum of both left and right lateral and medial orbitofrontal cortices (Morey et al., 2016), whereas the amygdala and accumbens were composed of the bilateral combination of ROIs for each subcortical region. The PCC consisted of the bilateral posterior dorsal and posterior ventral cingulate cortices (Holt et al., 2011).
For the ECS the LPFC was assembled by combining the bilateral dorsal and ventral sub-regions. The ventrolateral PFC comprised the pars opercularis, pars triangularis and pars orbitalis, whereas the dorsolateral PFC comprised the middle frontal gyrus and superior frontal gyrus (Vijayakumar et al., 2014). The posterior parietal cortex incorporated the bilateral inferior and superior parietal gyri (Desmurget et al., 2018; Sheppard et al., 2010), whereas the dACC was composed of the bilateral anterior and posterior middle cingulate (Demers et al., 2015). The VCC was assembled by combining the masks of both VS and ECS. All anatomical labels were extracted and merged using the FMRIB Software Library (Smith et al., 2004, https://fsl.fmrib.ox.ac.uk/fsl/).
The complete list of all anatomical ROIs used to delineate the VS and ECS are shown in Table 1, along with their central coordinates given in Montreal Neurological Institute (MNI) space.
Substance use measures
The following substance use measures were obtained from the Human Connectome Project Restricted Access Data and correlated with Valuation, Control and VCC average connectivity: Times Used Cocaine, Times Used Hallucinogens, Times Used Opiates, Times Used Sedatives, Times Used Stimulants and Times Used Marijuana. These substance use and frequency measures were collected using two approaches: a detailed questionnaire prepared specifically for the Human Connectome Project (for short-term alcohol and tobacco use), followed by the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA) for long-term alcohol, tobacco and illicit drug use (Bucholz et al., 1994).
Functional connectivity correlated with substance use behaviour
The rsfMRI signal was averaged over all voxels comprising each ROI (node) of the VS (12 nodes) and ECS (20 nodes), yielding a regionally-averaged signal for each node. The Pearson correlation coefficient in the regionally-averaged signals were then computed between all pairs of nodes, resulting in a 12 x 12 (VS), 20 x 20 (ECS) and 32 x 32 (VCC) connectivity matrix for each of the four scan runs (Rest 1-LR, Rest 1-RL, Rest 2-LR, Rest 2-RL). This computation was repeated for each subject. The four connectivity matrices computed from the four rsfMRI sessions for each subject were then averaged to improve the signal-to-noise ratio, yielding a single connectivity matrix for each of the three systems for every subject. These connectivity matrices were then averaged per subject, such that for each of the VS, ECS and VCC, each subject’s connectivity would be represented by a single value. Generalized linear models (GLM) were then computed to correlate the average connectivity of VS, ECS and VCC with each subject’s substance use profile. For each analysis, the substance measure in question (e.g. Cocaine Use) was used as a dependent variable, while network-averaged functional connection, use measures for all other substances (e.g. Drink Freq 7 Days, Tobacco Use Total 7 Days, Stimulant Use etc), gender and parental history of drug use were input as independent variables. Multiple comparisons were controlled using the False Discovery Rate (Benjamini and Hochberg, 1995). A full list of dependent and independent variables is shown in Table 2.
Results
Average connectivity of the VS, ECS and VCC is correlated, either independently or via a two-way interaction with substance use (Table 3). Both VS and ECS connectivity are positively correlated with Stimulant use, whereas both ECS and VCC connectivity are negatively correlated with Marijuana use. VS and VCC connectivities are negatively correlated with Sedative use.
VS, ECS and VCC connectivity is positively correlated with Cocaine use via a two-way interaction with Total Tobacco 7 Days (amount of tobacco consumed in week) but was negatively correlated via a two-way interaction with Sedative and Stimulants. VS, ECS and VCC connectivity is positively correlated with Marijuana use via a two-way interaction with Fathers’ Use History. ECS and VCC connectivity is negatively correlated with Cocaine use via a two-way interaction with Mothers’ Use History. VS, ECS and VCC connectivity is positively correlated with Opiate use via a two-way interaction with Tobacco Frequency 7 Days (frequency of tobacco use in a week). VS, ECS and VCC connectivity is positively correlated with Sedative use via a two-way interaction with Stimulant use, while VS and VCC connectivity is positively correlated with Sedative use via a two-way interactions with Alcohol Frequency 12 Months (frequency of consuming alcohol in a year) and Drinks/Day 12 Months (average number of drinks consumed per day in a year). VS and ECS connectivities are negatively correlated with Stimulant use via a two-interaction with Alcohol Frequency 12 Months.
VS connectivity is positively correlated with Cocaine use via a two-way interaction with Opiates and Hallucinogens, but negatively correlated via a two-way interaction with Marijuana. VS connectivity is positively correlated with Hallucinogen use via paired interactions with Cocaine and Sedatives but negatively correlated via interactions with Opiates and Total Tobacco 7 Days (total amount of tobacco consumed in a week). VS connectivity is positively correlated with Opiate use via two-way interaction with Cocaine. VS connectivity is also positively correlated with Sedative use via a two-way interaction with Hallucinogens. ECS connectivity is negatively correlated with Opiate use via a two-way interaction with Mothers’ Use History. ECS connectivity is negatively correlated with Stimulant use via a two-way interaction with Sedative use and Total Drinks 7 Days (total number of alcoholic drinks consumed in a week). Finally, VCC connectivity was negatively correlated with Stimulants via a two-way interaction with Mothers’ Use History.
Discussion
In this study we correlated the average connectivity of the Valuation and Control Systems, as well as the Valuation-Control Complex with drug use in the brains of healthy young adults. The positive correlation between VS and ECS with stimulants is related to sedative co-use (alcohol and sedatives), whereas the positive correlation between all three networks with sedative use occurs when network connectivity forms a two-way interaction with stimulants. We can thus infer cohort-specific motivations for using certain combinations of drugs based on trends observed in this study. Co-use pairs comprised of stimulating and sedating substances are popular due to their synergistic interactions. Stimulants energize the user, producing longer-lasting euphoria while off-setting sedating effects (Hernández-López et al., 2002), while sedatives and alcohol calm the user, ameliorating unpleasant side-effects of stimulant use such as anxiety and aggression (Fisk et al., 2011; Knackstedt and Ettenberg, 2005; Trujillo et al., 2011). On the other hand, the positive correlation between VS and VCC connections with sedatives via a two-way interaction with alcohol suggests a pain-killing combination. Co-use of sedating substances are commonly used for pain management (Jones et al., 2016; Peele, 2016; Vadivelu et al., 2014) especially when a single substance is found to be inadequate for the purpose (Weaver, 2015; Wilson et al., 2020).
The positive correlations observed among all three networks and cocaine use via two-way interactions with tobacco can be attributed to the concurrent use of both cocaine and tobacco (Marks et al., 2016), with the latter acting as a gateway to cocaine use (Kandel and Kandel, 2015). The positive correlation between VS connectivity and both cocaine as well as opiates could be indicative of “speedball” use. This cocaine-heroin mixture possess enhances feelings of euphoria an thus may act as a stronger reinforcer of cocaine-opiate combination use (Duvauchelle et al., 1998; Lacy et al., 2014). The VS acts as the middle ground for processing both pleasure and pain (Leknes and Tracey, 2008). Therefore, the positive correlation observed between VS, ECS and VCC connectivity and sedative substances (marijuana, sedatives) when paired with other sedating substances (such as alcohol) could indicate these combinations of substances are being used for pain management.
This study is not without limitations. There are various types of stimulants (Arria and Wish, 2006), sedatives (Ista et al., 2007; Lyphout et al., 2019) and hallucinogens (Skryabin et al., 2018), some of which have been manufactured to have pharmacological properties of other abusable substances (Klega and Keehbauch, 2018). Given that each drug has its own mechanism of action that target specific receptors or pathways, information regarding the specific types of drugs used, similar to the breakdown given for alcohol and tobacco use, may help shed additional light on the relationships between network connectivity and poly-drug use measures.
In conclusion, VS, ECS and VCC connectivity is correlated with poly-drug use, both independently and via interactions with other substances. A notable observation is that the connectivity of each network is able to distinguish between intake of different drugs, possibly along motivation and incentive lines (euphoria-seeking, pain-killing, etc). Reflecting on the relationships between substances, stimulant use was involved in the consumption of other drugs such as such as cocaine and sedatives. Alcohol and tobacco were linked also to poly-drug behaviour, correlating with all six illicit drugs analyzed. Although a preliminary study, these results open the door to more comprehensive analysis into the neuropsychiatric mechanics of poly-drug use, particularly the connections within the valuation system and valuation-control complex.
Data Availability
The date referred to in the manuscript was derived from the Human Connectome Project Healthy Young Adult Study
Funding
Kavinash Loganathan is grateful to IBRO-APRC and Yayasan Universiti Teknologi PETRONAS (YUTP) for funding to undertake this study. Eric Tatt Wei Ho gratefully acknowledges the Ministry of Higher Education, Malaysia for funding through the Higher Institution Center of Excellence (HI-CoE) program awarded to the Center for Intelligent Signal & Imaging Research, Universiti Teknologi PETRONAS and the YUTP-Fundamental Research Grant award. Vanessa Cropley was supported by an Australian National Health and Medical Research Council (NHMRC) Investigator Grant (1177370).
Acknowledgements
Kavinash Loganathan performed analysis, interpreted results and prepared this manuscript, Jinglei Lv was involved in experimental design, assisted in analysis and interpretation of results, Eric Tatt Wei Ho performed analysis and interpreted results, Vanessa Cropley assisted with brain mask generation and provided technical feedback, Andrew Zalesky was involved in experimental design, provided technical feedback and assisted with result interpretation.