Default mode network functional reorganization during early abstinence in polysubstance using emerging adults treated for opioid dependence

Abstract

We examined default mode network connectivity within the first 30 days of abstinence in emerging adults entering treatment for opioid dependence. There were significant associations between abstinence duration and coupling strength with brain regions within and outside of the network.

Introduction

Emerging adulthood describes the period from the late teens through the twenties during which brain neurodevelopment is ongoing[1]. In comparison to other age cohorts, emerging adults have the highest rates of opioid and other illicit substance use[2], are more likely to use opioids in combination with other substances[3], and have worse treatment outcomes[4]. Functional brain organizational differences among emerging adults are posited to play a role in their greater vulnerability to drug addiction and its negative cognitive neurological consequences, including lifetime executive function deficits[5].

Among opioid dependent patients, reduced regional coupling strength within the DMN has been observed in association with greater impulsivity[8], greater reported craving[9] and longer lifetime duration of heroin use[10]. These international studies, however, combine developmentally different age groups and exclude polydrug users, who comprise approximately 85% of persons entering opioid detoxification treatment in the United States[11]. No studies have examined the association between DMN connectivity and early abstinence (within the first 30 days), which is the timeframe when deficits in higher-order cognitive function are observed to peak in opioid dependence[12].

Based on evidence that reduced regional coupling strength within the DMN is associated with more lifetime opioid use and greater cognitive and clinical severity measures, we hypothesized that a longer duration of abstinence from all substances would be associated with stronger regional within network DMN connectivity in a sample of emerging adults entering treatment for opioid dependence.

Methods

Participants

Participants were recruited from the partial hospitalization program at McLean Hospital following completion of a 3–4 day inpatient detoxification program. Exclusion criteria included neurologic illness or injury, MRI contraindications, any psychiatric condition that would interfere with provision of consent or valid self-report, acute intoxication, substance withdrawal, pregnancy, or safety concerns. The McLean Hospital IRB approved this study. Informed consent was obtained after the study procedures had been fully explained.

Data collection and analyses

Substance use data in the past 30 days were collected using the Addiction Severity Index (ASI). The ASI was modified to include the question, “How many days has it been since you used [substance].” Data regarding DSM-IV substance use and co-occurring disorders were collected through structured clinical interview.

MRI scans were conducted using a Siemens Trio 3T scanner and 32-channel head coil. Structural MRI images were acquired for co-registration with the functional data using the following parameters: resolution=1.0 × 1.0 × 1.33 mm, repetition time (TR)= 2.1 s, echo time (TE)=3.3 ms, slices=128, matrix= 256 × 256, flip angle=7 degrees. Six-minute, eyes-open resting-state gradient echo-planar fMRI images were acquired with the following parameters: TR=2.5 s, TE=30 ms, flip angle=90 degrees, slices=42, voxel size=3.5 mm isotropic.

All data analyses were performed using the fMRI Software Library (FSL). Motion correction, brain extraction, slice timing correction, spatial smoothing with a Gaussian kernel of FWHM 6 mm, and a high-pass temporal filter with Gaussian-weighted least-squares, and straight-line fitting (100 s) pre-processing steps were conducted. Subject specific data were registered to the MNI152 2 mm³ standard space template. Functional MRI data were transformed using 2 × 2 × 2 mm resolution. The FSL MELODIC was used to remove sources of noise for each individual’s data.

A group-level Independent Component Analysis (ICA) using FSL MELODIC was then performed to define the DMN for the subject sample. Consistent with our prior work[13–14], dimensionality was fixed to 35 components to examine large-scale resting-state networks. The DMN was identified visually and its spatial distribution matched prior work (15). To calculate subject-specific time courses and spatial maps for the DMN, we used the dual regression approach implemented using FSL.

To assess if DMN coupling varies relative to abstinence duration the non-parametric permutation method (FSL Randomise) was used to correlate days since last substance use with DMN coupling (5,000 permutations). Cluster-based thresholding was corrected for multiple comparisons by using the null distribution of the max cluster size across the image (cluster-corrected z=2.3, p < 0.05,).

Results

Thirteen Caucasian emerging adults (8 male, 5 female) between 18–27 years old completed their fMRI brain scan within 24-hours of partial hospital admission for treatment for opioid dependence. Equal numbers (46%) presented for heroin or prescription opioid dependence, and 8% for both. 15% were additionally receiving opioid agonist therapy, which was continued during and following detoxification treatment. Subjects reported a Mean (SD) of 1(±1) year of lifetime opioid use, and 5 (±6) days since using any psychoactive drug (including opioids, cannabis, cocaine, sedatives, and alcohol; and excluding prescribed opioid agonist medication and nicotine). There were no significant relationships between the number of days since last substance use (range 1–23 days) and age (r=−.20, p=.50), sex (r=−.16, p=.61), or education (r=.41, p=.19). Subjects met DSM-IV dependence criteria for the following additional substances (54% cannabis, 23% sedatives, 15% cocaine, 8% alcohol), and met criteria for the following co-occurring disorders (23% substance-induced mood or anxiety disorder, 15% depression, 7% panic disorder). Spontaneous resting-state BOLD signal fluctuations ascribable to the DMN were detected in our sample (see Figure 1A); and there was a significant association between abstinence duration and strength of connectivity between the DMN and brain regions identified in Figure 1B.

FIGURE 1.

A. Default Mode Network (DMN). The DMN encompasses a set of distributed brain regions including: (a) dorsomedial prefrontal cortex, (b) ventromedial prefrontal cortex, (c) ventral precuneus cortex, (d) posterior cingulate cortex, (e) inferior parietal cortex, (f) lateral temporal cortex and (g) hippocampal formation. Using FSL, an ICA was conducted to identify the DMN. Spontaneous resting-state BOLD signal fluctuations ascribable to the DMN were detected in our sample. These data were visually inspected to ensure component singularity and spatial consistency with previously published work[15]. B. DMN Regional Coupling and Abstinence. Red overlay represents brain regions with significant coupling with the DMN (green underlay) relative to increased abstinence duration. Brain regions with significant within-network DMN coupling are identified (a–d) in black lettering, and include: (a) right precuneus; (b) posterior cingulate cortex, which extends into (c) retrosplenial cortex (forming a posterior cingulate/retrosplenium cluster); and (d) hippocampal formation. Brain regions with significant outside-of-network coupling are identified (e–i) in white lettering, and include: (e) right insula, which extends into (f) putamen (forming a putamen/insula cluster); (g) thalamus; (h) bilateral lateral occipital cortex; and (i) bilateral cerebellum. Cluster corrected z = 2.3, p < 0.05, 5,000 permutations. Coordinates are in (mm) in Montreal Neurological Institute (MNI-152) standard space.

Discussion

Consistent with our hypothesis, there was a significant positive relationship between abstinence duration and DMN strength of within-network coupling. Our findings are similar to those from the only other study to examine the association between strength of DMN coupling and drug use patterns in opioid dependence treatment patients[8]. In that study, within-network DMN coupling was associated with duration of lifetime opioid use; and in our study it was associated with recent (past 30 day) substance use. Other authors have speculated that a longer duration of alcohol sobriety is associated with increased within-network DMN coupling because it provides time for recovery and repair[11]. This may be relevant to the present findings in that DMN connectivity changes within the first 30 days of abstinence may also reflect recovery and repair.

In addition to our hypothesized within-network changes, we also found strength of diffuse outside-of-network connectivity was positively associated with abstinence duration. We can’t draw conclusions about this incidental finding within the constraints of our small pilot study, but other studies may help identify relevant variables and questions for future studies. For example, in a study that examined resting state fMRI data via topographical analysis, polysubstance users in treatment for cocaine dependence had stronger and more diffuse but less efficient inter-regional resting-state network connections in comparison to control subjects[12]. These authors speculated that connection strength might increase to compensate for a “hyperconnected” and less efficient resting state network in the addicted brain, but it is unknown how abstinence duration may affect these relationships.

This pilot study has several limitations, including a small sample size, absence of data from a second time point, reliance on self-report for substance use data, absence of data regarding nicotine use recency and its unknown effects on the DMN, and inability to control for multiple substances. However, this study also has a high degree of ecological validity and is highly relevant to the management of persons entering opioid dependence treatment who have the highest rates of opioid and co-occurring substance use, and the poorest opioid dependence treatment outcomes. Future studies should examine DMN coupling as a repeated measure, and with a comparison group and larger sample size to better understand how time and abstinence duration affect within and outside-of-network coupling.