Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 May 1.
Published in final edited form as: Alcohol Clin Exp Res. 2019 Mar 29;43(5):812–821. doi: 10.1111/acer.13998

The Relationship between Regional CBF Estimates and Alcohol Problems at 5-Year Follow-Up: The Role of Level of Response

Kelly E Courtney 1, M Alejandra Infante 1, Gregory G Brown 1,2, Susan F Tapert 1, Alan N Simmons 1,2, Tom L Smith 1, Marc A Schuckit 1,*
PMCID: PMC6502665  NIHMSID: NIHMS1016764  PMID: 30924954

Abstract

Background:

Acute alcohol consumption is associated with temporarily increased regional cerebral blood flow (CBF). The extent of this increase appears to be moderated by individual differences in the level of response (LR) to alcohol’s subjective effects. The low LR phenotype is a known risk factor for the development of alcohol problems. This study investigates how the low LR phenotype relates to the relationship between alcohol-related changes in CBF and alcohol problems 5 years later.

Methods:

Young adults (ages 18–25) were selected based on their LR to alcohol and underwent a neuroimaging protocol including arterial spin labeling and functional scans. These participants were re-contacted ~5 years later and assessed on alcohol outcomes. A final sample of 107 subjects (54 low and 53 high LR subjects) was included in the analyses. Whole-brain analysis revealed five clusters of significant alcohol-induced, versus placebo-induced, CBF changes that were consistent with a previous report. Peak alcohol – placebo CBF response was extracted from these regions and, along with LR group, submitted to a hierarchical linear regression predicting alcohol problems. Analyses controlled for age, sex, and baseline alcohol problems.

Results:

In the regression analysis, greater alcohol – placebo CBF difference in the right middle/superior/inferior frontal gyri and bilateral anterior cingulate gyri clusters predicted greater future alcohol problems for the low LR group, whereas this relationship was not found to be significant in the high LR group.

Conclusions:

This study demonstrates a clinically important relationship between CBF and future alcohol problems, particularly in individuals with a low LR phenotype. These initial results help to elucidate the neurobiological pathways involved in the development of alcohol use disorders for individuals with low LR.

Keywords: cerebral blood flow, level of response, alcohol problems, arterial spin labeling, young adults

Introduction

Alcohol use disorders (AUDs) remain a highly prevalent and burdensome condition for the affected individuals and society. According to the recent National Survey of Drug Use and Health results, approximately 6% of individuals ages 12 or older, or 11% of young adults (ages 18–25 years old), met criteria for an alcohol use disorder in 2016 (Substance Abuse and Mental Health Services Administration, 2017). The excessive use of alcohol is associated with a number of negative physical and psychosocial consequences (Piano et al., 2017, Squeglia et al., 2014, Gmel and Rehm, 2003), together resulting in approximately $249 billion in costs to the U.S. in 2010 (Sacks et al., 2015). These substantial estimates underscore the need for additional research on the origins of AUDs to facilitate targeted prevention and intervention strategies.

The development of an AUD is likely to involve a number of genetic and environmental influences. Further, the traditional conceptualization of AUD includes a diverse set of individual phenotypes, leading to significant heterogeneity in disorder development and presentation (Hyman, 1999). To gain traction in AUD research, many researchers have turned to the study of more narrowly defined intermediate phenotypes related to AUDs (Claus et al., 2011, Hines et al., 2005). The use of an intermediate/endophenotype approach has led to enhanced power in genetic (Dick et al., 2006) and neurobiological studies (Ray et al., 2010), and has provided greater opportunity for the discovery of early points of intervention in those at high risk for disorder development (Savage et al., 2015, Schuckit et al., 2016).

Intermediate phenotypes that capture how an individual responds to alcohol have proven to be clinically important and reliable risk factors for AUD development (King et al., 2014, Schuckit et al., 2007). In particular, the low level of response (low LR) phenotype has been shown to be a genetically influenced characteristic (Schuckit, 2009) that predicts future heavy drinking and risk for developing AUDs (Chung and Martin, 2009). This model suggests that individuals with a low LR profile, meaning they need higher blood alcohol concentrations (BACs) to experience alcohol effects, are at greater risk for heavy drinking and alcohol problems than individuals who exhibit a high LR profile (Schuckit et al., 2007, Zhang et al., 2015, Schuckit et al., 2008). The low LR profile has been associated with a general dampening of physiological markers in these individuals following alcohol administration, including reduced alcohol-related effects on adrenocorticotropic hormone and prolactin release and less intense changes to cortical electroencephalograms and event-related potentials, as compared to their high LR counterparts (Ehlers and Schuckit, 1988, Schuckit et al., 1987, Schuckit et al., 1988b, Schuckit et al., 1988a).

An individuals’ LR also appears to have measurable neural correlates. Functional magnetic resonance imaging (fMRI) studies using cognitive and emotional-based tasks to probe the low LR phenotype suggest two primary findings despite similar task-related behavioral performance: (1) low LR individuals, as compared to high LR individuals, tend to have greater regional blood-oxygen-level dependent (BOLD) response under baseline or placebo conditions(Schuckit et al., 2012, Trim et al., 2010, Tapert et al., 2004), potentially reflecting the need for greater cognitive effort during task performance (Schuckit et al., 2016), and (2) low LR individuals show a general tendency for dampened or attenuated BOLD response following acute alcohol administration, as compared to high LR individuals who often show the opposite response post-alcohol (Paulus et al., 2012, Schuckit et al., 2012). Importantly, LR-related BOLD response in the right middle frontal gyrus and the left anterior insula during an emotion processing task was found to predict future alcohol consumption and alcohol-related problems at 5 year follow-up (Schuckit et al., 2016), suggesting the presence of clinically relevant variation in the neural profile of the LR phenotype.

The study of cerebral blood flow (CBF) could provide additional neurobiological information regarding the heightened risk for heavier drinking, alcohol problems and AUDs associated with the low LR profile. CBF is tightly coupled with glucose metabolism (Jueptner and Weiller, 1995) and brain function (Raichle et al., 1976), and reductions in regional CBF and glucose metabolism in the sober state, particularly in frontal brain regions, have been observed with the misuse of multiple substances (Murray et al., 2015), including alcohol (Moselhy et al., 2001, Hamdi et al., 2003). These regional CBF reductions have been associated with severity of AUD (Hamdi et al., 2003) and important functional outcomes, such as relapse to drinking after treatment (Durazzo et al., 2010) and neurocognitive functioning (Goldstein et al., 2004).

Although there is general agreement that acute alcohol consumption at low to moderate levels is associated with temporarily increased regional CBF (Gundersen et al., 2013, Khalili-Mahani et al., 2011, Tiihonen et al., 1994), the intensity of this increase appears to be moderated, in part, by individual differences in LR (Tolentino et al., 2011, Strang et al., 2015). Consistent with a general dampening of alcohol-induced physiological effects, non-dependent young adult alcohol drinkers with a low LR profile demonstrate less increase in CBF after acute alcohol consumption as compared to those with a high LR, particularly in frontal and cingulate brain regions.

Taken together, the literature suggests a notable relationship between frontal CBF, alcohol-related outcomes, and LR. However, it remains unknown whether the increase in CBF following acute alcohol administration is clinically meaningful as a potential biologically based marker of future alcohol problems and whether the known differences in LR-related CBF increases following alcohol affect this potentially predictive relationship. The current paper attempts to address these research questions by testing whether alcohol – placebo-related CBF estimates can predict alcohol problem outcomes 5-years later in a longitudinal sample of young adults who differ in their LR to alcohol. Analyses focused on five frontal clusters associated with alcohol-related CBF changes first identified in a data-driven, whole-brain, voxel-wise analysis by Tolentino et al. (2011) which included a subset of the present study data. We hypothesized that the magnitude of change in these regional CBF estimates during alcohol – placebo challenge at baseline would predict alcohol problems at 5-year follow-up (i.e., less CBF change to alcohol will predict more alcohol problems later). Given the increased risk associated with the low LR phenotype, we further hypothesized that the predictive relationship of the magnitude of CBF change will be stronger in low LR than high LR. Given the novelty of the study aims, no cluster specific hypotheses were made.

Materials and Methods

Participants and Study Design

The data for the analyses were taken from a recently completed study (see Schuckit et al., 2016 for a more thorough description of methods). Briefly, potential participants were initially identified through mailings to students at a local university, using procedures approved by the UCSD Human Research Protections Program. The mailings included questions about demography and the Self-Report of the Effects of alcohol (SRE) questionnaire, which includes 4 questions pertaining to the number of standard alcoholic drinks (~10 g ethanol) needed to experience a range of effects (i.e., feeling effects, slurred speech, unsteady gait, and unwanted falling asleep) the first 5 times of drinking (before chronic tolerance was likely to have developed) (Schuckit et al., 1997).

Consistent with all our prior studies of LR and most additional studies of this phenotype, young adults with more limited drinking histories (e.g., no histories of alcohol use disorders) were targeted to minimize the chances that we are measuring the results of consistent heavy drinking and to determine if these early responses to alcohol can serve as predictors of future alcohol use behaviors. Thus, inclusion criteria for the study were: (1) ages 18 to 25 and (2) having consumed at least one full standard alcoholic drink in the past (due to the requirement of some previous exposure to alcohol for completion of the SRE and alcohol challenge procedures). Exclusionary criteria were: (1) a lifetime or current diagnosis of alcohol or other drug dependence, bipolar or schizophrenia (American Psychiatric Association, 1994); (2) a current medical condition or use of a medication that might interfere with an alcohol challenge or brain blood flow; (3) previous head trauma with loss of consciousness >3 minutes; (4) current pregnancy; (5) left handedness; and (6) MRI contraindications (e.g., claustrophobia, irremovable metal).

Participants were pre-selected based on their SRE scores (~upper and lower SRE thirds) and scheduled for a face-to-face interview which included the full Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA) interview (Schuckit et al., 2012, Bucholz et al., 1994). The SSAGA was designed to assess psychological, physical, and social manifestations of alcohol abuse/dependence and other psychiatric disorders and includes questions on recent alcohol consumption (quantity and frequency of current alcohol use) and alcohol-related problems. If deemed eligible, participants were then invited to the laboratory to complete an acute alcohol challenge procedure to verify their LR profile. The alcohol challenge procedure included the consumption of 0.70 ml ⁄ kg for women; 0.75 ml ⁄ kg for men of laboratory-grade ethanol consumed over ~10 minutes (resulting peak BACs of ~60 mg/dl). Participants then completed the 7-item version of the Subjective High Assessment Scale (SHAS), an analogue self-report measure of positive and negative subjective feelings of alcohol intoxication (Eng et al., 2005), every 15–30 minutes for ~3 hours. Approximately 80% of the pre-selected participants displayed consistent LR profiles on the SHAS and SRE (see Trim et al., 2010 for more detailed methods).

For the baseline imaging protocol, 120 participants were identified (60 low LR and 60 high LR) to take part in two neuroimaging sessions administered in random order, one after consuming the .70 to .75 ml/kg dose of ethanol as described above and one after a no-alcohol placebo beverage, each consumed over 10 minutes. Participants were also assessed again on indices of recent alcohol consumption and alcohol problems/alcohol use disorder symptoms using selected questions from the SSAGA interview (Schuckit et al., 2012, Bucholz et al., 1994). Possible alcohol problems included the 11 DSM-IV abuse and dependence items plus blackouts, morning drinking, efforts to stop drinking, self-help group attendance, and receiving professional help for alcohol-related difficulties (total possible problem count of 16). The major focus for the baseline results of the original papers was on the relationship of the LR status of the subject to the BOLD response contrast patterns in relevant regions of interest, with arterial spin labeling (ASL) measures of CBF used as a covariate (Paulus et al., 2012, Schuckit et al., 2012, Trim et al., 2010).

Participants were then re-contacted by telephone approximately 5 years later and assessed again on indices of recent alcohol consumption (quantity and frequency of alcohol use during follow-up) and alcohol problems/alcohol use disorder symptoms using the selected SSAGA questions. Complete CBF and follow-up data are available for 109 subjects (55 low and 54 high LR participants).

Neuroimaging Protocol

Imaging data were collected at the UCSD fMRI Center with a 3-Tesla General Electric (Milwaukee, WI) Signa Excite HD scanner using an 8-channel head array coil. Scanning began 22 minutes after the start of beverage administration and included a number of functional scans (see Paulus et al., 2012, Schuckit et al., 2012, Trim et al., 2010) prior to the resting ASL scan. Collection of resting-state ASL data began approximately 60 minutes after the start of beverage consumption, thus occurring at peak or the initial phase of the descending BAC limb. Each scan session included a sagittally acquired high-resolution spoiled gradient-recalled (SPGR) anatomical sequence (25 cm field of view; 256 × 256 matrix; 124 slices each 1.0 mm thick covering the whole brain; 4.8 ms echo time; and 20 ms repetition time). Three sequences were acquired to obtain absolute CBF measurements during each session. Resting brain blood perfusion was measured with pulsed ASL using a modified flow-sensitive alternating inversion recovery sequence with both presaturation pulses and PICORE QUIPSS 2 postinversion saturation pulses and a spiral readout with 4 interleaves to reduce signal dropout (22 × 22 cm field of view, a 64 × 64 matrix, 3.2 ms echo time, 2500 ms repetition time, postsaturation and inversion times of TI1 = 600 ms and TI2 = 1600 ms, tag thickness 10 cm, tag to proximal slice gap 1 cm, 20 5 mm axial slices, and 40 volumes for 20 tag + control image pairs) (Wong, 2005). In addition, a scan without inversion pulses was acquired to obtain an estimate of the equilibrium magnetization of cerebral spinal fluid (CSF), and a minimum contrast image was acquired to adjust for coil inhomogeneities (Restom et al., 2007).

Data Processing

Imaging data were processed using Analysis of Functional NeuroImages (AFNI; afni.nimh.nih.gov; Cox, 1996), FMRIB Software Library (FSL, Oxford, United Kingdom) (Smith et al., 2004), and locally created MatLab scripts. As in the original study (Tolentino et al., 2011), each ASL dataset was first reconstructed using the SENSE algorithm (Pruessmann et al., 1999, Weiger et al., 2002) to reduce sensitivity to the modulations that occur between shots caused by physiological fluctuations or motion. Second, an automated MatLab script including AFNI and FSL tools processed the ASL data. Third, the ASL time series were co-registered to the middle time point to minimize the effects of participant motion. Fourth, surround subtraction of the tag – control time series was performed to create an uncorrected perfusion time series and slice timing delays were accounted for, making the inversion time (TI2) slice specific.

Each participant’s session specific high-resolution T1-weighted image was skull stripped using AFNI’s 3dSkullStrip. A mask from the same session average A3 or ASL image was then calculated using 3dAutomask and applied to the CBF data. The high-resolution anatomicals were then aligned to these masked CBF images using AFNI’s align_epi_anat script and segmented using FSL’s FAST algorithm to define CSF, gray matter (GM) and white matter (WM) regions. The aligned high resolution and full-field CBF images were warped to Talairach space using AFNI’s auto_tlrc function and resampled to a 4 × 4 × 4 mm resolution grid with AFNI’s adwarp. Voxels with negative intensities were replaced with zero (Brown et al., 2003). Data from both the alcohol and placebo sessions were visually screened for data quality and alignment.

Statistical Analyses

Whole brain, voxel-level analysis was conducted using a paired t-test (AFNI 3dttest++) to contrast perfusion values between the placebo and alcohol sessions (alcohol – placebo and placebo – alcohol contrasts computed). To control for Type I error, the Clustsim nonparametric randomization/permutation option of 3dttest++ was used with a conservative voxel-wise alpha of 0.001 and cluster-wise alpha of 0.05, which estimated a cluster size threshold of 7 contiguous voxels. The AFNI version used for all analyses included the bug-correction in 3dClustSim. The randomization step of the clustering process randomizes the signs of the model residuals, whereas the permutation step permutes the imaging volume assignment across the two image sets in order to determine a 0.05 cluster size threshold. As discussed by Eklund et al. (2016), nonparametric permutation tests offer precise control of false positives. Furthermore, AFNI’s 3dtest++ randomization step alone has been shown to produce false positive rates compatible with the nominal 95% confidence interval (FPR: 3.65–6.35%; Cox et al., 2017). To be consistent with the original manuscript (Tolentino et al., 2011), an averaged grey matter mask was created from the individual FAST grey matter volumes using AFNI’s 3dcalc function, which was then used to restrict the results of the t-test. Two subjects (one high LR, one low LR) exhibited extreme alcohol – placebo CBF estimates across all clusters of interest (>50 ml ⁄ 100 g ⁄ min difference between conditions) and were excluded from the analyses, leaving a final sample of 107 subjects (54 low and 53 high LR subjects).

Five clusters of interest corresponding to the original article results (Tolentino et al., 2011) were then identified from the significant clusters of whole-brain alcohol – placebo contrast activation. CBF estimates from each of these clusters were extracted from each subject. Independent samples t-tests were conducted to compare high and low LR groups on these five alcohol – placebo CBF cluster estimates.

To test the predictive utility of the CBF estimates, the mean alcohol – placebo activation for these five clusters were entered into a hierarchical linear regression model predicting alcohol problem count at follow-up. Participant sex was controlled for given that females tend to exhibit greater increases in global perfusion following acute alcohol administration (Marxen et al., 2014). Thus, sex, age, and baseline level of alcohol problems were entered in Step 1 of the model as they represented theoretically important a-priori covariates. LR group membership and the five main effect cluster CBF estimates were entered in Step 2, followed by the interaction between the cluster CBF estimates and LR group entered in Step 3. All values were entered in their original metric (un-centered) given the meaningfulness of 0 in the CBF cluster estimates (0 representing no difference between alcohol and placebo) and the LR group variable (low LR coded as 0, high LR coded as 1) (Shieh, 2011). An exploratory whole-brain correlation with follow-up alcohol problems count including LR group as a between subjects factor was also conducted on alcohol – placebo CBF difference maps using AFNI 3dttest++ (p < .01, uncorrected). In addition, an exploratory hierarchical linear regression model consistent with the model above was ran with average alcohol use frequency per month serving as the baseline covariate and outcome of interest.

Results

Participants

The sample demography and alcohol use characteristics relevant to the key analyses are presented in Table 1. As shown in this table, participants were approximately 20 and 25 years old at baseline and follow-up, respectively. There were roughly equal numbers of males and females in the sample. As expected, the low LR group generally reported greater maximum drinks and alcohol problems than the high LR group both at the baseline and follow-up time points; yet, only the baseline differences met statistical threshold (p < .05). Both LR groups exhibited significantly greater maximum drinks and alcohol problems at follow-up as compared to baseline (p values < .001). No differences were observed between baseline and follow-up time points with respect to usual drinks in either group (p values > .28).

Table 1.

Sample demographics.

Variable % Frequency or Mean (SD) Baseline LR Group Difference
(p values)		 Follow-Up LR Group Difference
(p values)
Low LR
(n = 54)		 High LR
(n = 53)
Baseline Follow-Up Baseline Follow-Up
Age 19.72 (1.42) 25.02 (1.71) 19.92 (1.49) 25.09 (1.76) .47 .82
% male 50.0 45.3 .62 .62
Ethnicity (%) .22 .22
 - Hispanic 20.4 32.1
 - Non-Hispanic 77.8 66.0
 - Unknown/other 1.8 1.9
Usual drinks per drinking day 3.92 (1.70) 3.56 (1.77) 3.16 (1.78) 3.19 (1.44) .03 .24
Max drinks per drinking day 7.94 (3.89) 11.09*** (4.50) 5.89 (3.30) 9.57*** (4.03) <.01 .07
Average drinking days per month 7.18 (4.56) 11.98*** (6.68) 6.04 (4.87) 10.23*** (7.16) .25 .22
Alcohol problems count 0.80 (0.92) 2.96*** (2.63) 0.38 (0.60) 2.09*** (2.18) <.01 .07
Open in a new tab

Note:

***

indicates significant change from baseline to follow-up (p values < .001).

Neuroimaging

Whole-brain.

Given our expanded sample from 88 subjects in the prior paper (Tolentino et al., 2011) to 107 in the current paper, a whole-brain alcohol – placebo CBF contrast analysis was conducted in order to replicate and extend upon the original findings by Tolentino and colleagues (2011). The alcohol – placebo contrast revealed significant alcohol-related CBF response in 23 clusters (see Table 2). These results are largely consistent with the original analysis despite the inclusion of 19 additional subjects and more stringent statistical thresholding practices. Also consistent with the original analysis, no significant results were obtained for the reverse contrast, placebo – alcohol.

Table 2.

Clusters of significant CBF response from the whole-brain alcohol – placebo contrast across all subjects (thresholded at a voxel-wise alpha of 0.001 and clusterwise alpha of 0.05; ≥ 7 contiguous voxels). Bolded clusters represent clusters consistent with the previous report (Tolentino et al., 2011) and were extracted for follow-up analyses.

Cluster Location Voxels X Y Z
1 R middle/superior/inferior frontal gyri 106 46 43 8
2 L middle/superior frontal gyri 70 −34 39 40
3 Bil medial/superior frontal and cingulate gyri 40 −6 −9 64
4 Bil thalamus 39 −6 −21 8
5 Bil anterior cingulate gyri 35 2 39 0
6 R inferior temporal gyrus/fusiform gyrus 28 62 −17 −24
7 R precentral gyrus 21 50 −5 48
8 R superior/medial frontal gyri 19 2 −9 64
9 R inferior frontal gyrus/precentral gyrus 16 50 15 8
10 R inferior parietal lobule/insula 15 58 −33 20
11 L inferior temporal gyrus/fusiform gyrus/culmen 14 −58 −49 −12
12 L inferior/middle temporal gyrus 13 −62 −25 −16
13 Bil cingulate/precuneus 10 2 −61 28
14 L supramarginal gyrus/inferior parietal lobule 10 −58 −49 32
15 L inferior temporal gyrus/fusiform gyrus 9 −58 −9 −24
16 L middle temporal gyrus 9 −62 −49 0
17 R middle/superior frontal gyrus 9 42 15 52
18 L precentral gyrus/middle frontal gyrus 9 −42 −1 48
19 R middle/superior temporal gyrus 8 66 −21 −12
20 R middle temporal gyrus 8 62 −37 0
21 L parahippocampal gyrus 7 −26 −37 4
22 L middle frontal gyrus 7 −30 51 −4
23 R supramarginal gyrus/inferior parietal lobule 7 58 −53 36
Open in a new tab

Note: X, Y, and Z Talairach coordinates indicate the location of peak voxel activation within each cluster. L = left, R = right, Bil = bilateral.

Cluster identification.

Of the whole-brain results, five clusters largely consistent with the original analysis by Tolentino and colleagues (2011) were selected for further analysis (presented in bold in Table 2). As depicted in Figure 1, these clusters included the: (1) right middle/superior/inferior frontal gyri (red), (2) left middle/superior frontal gyri (purple), (3) bilateral medial/superior frontal and cingulate gyri (green), (4) bilateral anterior cingulate gyri (yellow), and (5) right precentral gyrus (blue).

Figure 1.

Figure 1.

Open in a new tab

Clusters of significant alcohol – placebo CBF response extracted for prediction analyses.

LR differences in alcohol-related CBF.

In an effort to replicate the original study findings and to determine whether the LR groups differed in alcohol-related CBF response within the selected clusters, mean CBF responses to alcohol – placebo were extracted from the five clusters identified and statistically compared across groups. Consistent with the original paper, the low LR group generally exhibited a smaller CBF change in response to alcohol – placebo administration across the extracted clusters as compared to the high LR group; however, only the bilateral medial/superior frontal and cingulate gyri cluster reached statistical significance (t (1,105) = −2.398, p = .018; yellow in graph; see Figure 2).

Figure 2.

Figure 2.

Open in a new tab

Mean CBF values by response group for each cluster extracted.

CBF Cluster Prediction of Alcohol Problems at Follow-Up

Next, a hierarchical linear regression model was conducted to determine the predictive utility of the CBF response, and the possible moderating role of LR group, on alcohol problems at 5-year follow-up. The results of this analysis are depicted in Table 3. Step 1 evaluates how three baseline characteristics used as covariates relate to the number of alcohol problems during the follow-up period, with results indicating significant contribution for only baseline alcohol problems and the three items explaining 8% of the variance (the R2) (F [3,103] = 2.86, p = 0.04).

Table 3.

Summary of the hierarchical regression analysis results predicting follow-up alcohol problems.

Step 1 Step 2 Step 3
Model Variables B (SE) B (SE) B (SE)
Step 1 – Demographic Covariates
Sex 0.61 (0.47) 0.44 (0.48) −0.08 (0.49)
Age −0.13 (0.16) −0.05 (0.17) −0.02 (0.17)
Baseline Alcohol Problems 0.82** (0.29) 0.55† (0.32) 0.40 (0.32)
Step 2 – Main/Conditional Effects
LR Group −0.43 (0.50) 0.93 (0.64)
Cluster 1: R middle/superior/inferior frontal gyri 0.01 (0.04) 0.09† (0.05)
Cluster 2: L middle/superior frontal gyri 0.02 (0.04) −0.04 (0.05)
Cluster 3: Bil medial/superior frontal and cingulate gyri −0.06† (0.03) −0.08† (0.04)
Cluster 4: Bil anterior cingulate gyri 0.03 (0.02) 0.09* (0.04)
Cluster 5: R precentral gyrus 0.01 (0.03) −0.01 (0.03)
Step 3 - Interactions
Cluster 1 × Group −0.16* (0.07)
Cluster 2 × Group 0.10 (0.07)
Cluster 3 × Group −0.01 (0.06)
Cluster 4 × Group −0.10* (0.05)
Cluster 5 × Group 0.04 (0.05)
R2 0.08 0.13 0.24
R2 Change 0.08 0.05 0.11
F for change in R2 2.86* 1.01 2.72*
Open in a new tab

Note:

**

p < .01,

*

p < .05,

†

p < .10.

Step 2 evaluated whether there were significant main effects in the regression analysis when it was expanded to include LR group and the relevant CBF clusters. This step revealed no significant main effects for the key variables entered into these analyses, and the 5% increase in R2 from Step 1 to Step 2 did not account for a significant increase in variance over Step 1 (F change [6,97] = 1.01, p = 0.42).

Step 3 extends Steps 1 and 2 by adding all interactions between LR group and CBF cluster estimates. This step increased the R2 to 0.24, with a significant R2 change of 0.11 (F[5,92] = 2.716, p < 0.03), and a total of 24% variance in predicting follow-up alcohol problems explained. Within this third step, controlling for sex, age, and baseline alcohol problems, a significant interaction was observed between LR group and alcohol – placebo CBF response in the right middle/superior/inferior frontal gyri cluster (Cluster 1; B = −0.16, p = 0.04). Examination of each group’s conditional effects (i.e., the effect of each group separately, controlling for the other model variables) of the alcohol – placebo CBF response in the right middle/superior/inferior frontal gyri cluster revealed that greater CBF response in this cluster predicted greater alcohol problems for the low LR group at a trend level (B = 0.09, p = 0.09), while greater alcohol – placebo CBF response in this cluster was non-significantly associated with less alcohol problems for the high LR group (model reran with high LR group recoded as 0: B = −0.07, p = 0.13).

A similar significant LR group interaction with alcohol – placebo CBF response was observed for the bilateral anterior cingulate gyri cluster (Cluster 4; B = −0.10, p = 0.03). Examination of the conditional effects revealed that greater alcohol – placebo CBF response in this cluster significantly predicted greater alcohol problems for the low LR group (B = 0.09, p = 0.02), while greater alcohol – placebo CBF response in this cluster was non-significantly associated with less alcohol problems for the high LR group (model reran with high LR group recoded as 0: B = −0.01, p = 0.74).

These data are also presented in Figures 3 and 4. Figure 3 depicts the interaction between LR group and alcohol – placebo CBF response from the right middle/superior/inferior frontal gyri cluster (Cluster 1), whereas Figure 4 depicts the interaction between LR group and alcohol – placebo CBF response from the bilateral anterior cingulate gyri cluster (Cluster 4). In both figures, LR moderated the predictive relationship between CBF response to alcohol and alcohol problems at follow-up such that greater alcohol-related CBF response (plotted on the x-axes) generally predicted greater alcohol problems (plotted on the y-axes) for the low LR group (shown in blue), yet CBF response and alcohol problems were not significantly related in the high LR group (shown in red).

Figure 3.

Figure 3.

Open in a new tab

Scatter plot depicting the interactive effects between alcohol – placebo cerebral blood flow (CBF) values extracted from the right middle/superior/inferior frontal gyri cluster (x-axis) and level of response (LR) group on the predicted alcohol problem count scores estimated from the hierarchical analysis (y-axis).

Figure 4.

Figure 4.

Open in a new tab

Scatter plot depicting the interactive effects between alcohol – placebo cerebral blood flow (CBF) values extracted from the bilateral anterior cingulate gyri cluster (x-axis) and level of response (LR) group on the predicted alcohol problem count scores estimated from the hierarchical analysis (y-axis).

Exploratory whole-brain correlations (voxel p<.01, no cluster correction) revealed activation largely constrained within the clusters of interest and were consistent with the results of the regression analysis. The hierarchical model predicting average alcohol use frequency during follow-up did not reveal any significant main or interaction effects pertaining to the variables of interest.

Discussion

The goal of the present analyses was to investigate the ability of regional alcohol-related cerebral blood flow (CBF) response to predict alcohol problems 5-years later within a longitudinal sample of young adults who differ in their level of response (LR) to alcohol. The primary finding was that greater alcohol – placebo CBF contrast extracted from the right middle/superior/inferior frontal gyri and bilateral anterior cingulate gyri clusters predicted greater future alcohol problems for the low LR group, whereas alcohol-related CBF response in these clusters and future alcohol problems were not significantly related in the high LR group. These findings were observed even after controlling for the impact of sex, age, and the number of alcohol problems at baseline. Importantly, the addition of LR group and alcohol-related CBF response interactions to the statistical model accounted for a significant amount of variance in later alcohol problems (24%) such that prediction of outcomes by CBF was only seen in the low LR group, suggesting the potential importance of considering the combination of both alcohol-related CBF and LR in predicting future alcohol problem development.

With respect to our first hypothesis, less alcohol-related CBF change in bilateral medial/superior frontal and cingulate gyri was associated with more alcohol problems at follow-up, regardless of LR; however, this effect did not exceed the statistical threshold. Instead, consistent with our second hypothesis, the relationship between alcohol-related CBF response and later alcohol problems was only statistically significant within individuals with low LR to alcohol. In line with an overall dampening of alcohol-induced physiological effects (Ehlers and Schuckit, 1988, Schuckit et al., 1987, Schuckit et al., 1988b, Schuckit et al., 1988a) and previous CBF analyses on LR (Tolentino et al., 2011, Strang et al., 2015), the low LR group in the present study also generally exhibited a smaller alcohol-related CBF response across the brain as compared to the high LR group. In other words, despite a general tendency for lower alcohol – placebo CBF response in the low LR versus high LR groups, within the low LR group, individuals with greater alcohol-related CBF responses evinced greater alcohol problems at follow-up. Taken together, these results suggest that the combination of a low LR profile and increased alcohol-related CBF response represents an especially risky phenotype for the development of later alcohol problems.

Consistent with the original baseline analysis on a subset of these subjects (Tolentino et al., 2011), the largest clusters of alcohol – placebo CBF response in the whole-brain analysis were observed in the frontal and anterior cingulate gyri. The frontal systems are known to be affected by acute alcohol administration (Zheng et al., 2015) and BOLD activation in the right middle frontal gyrus during an emotional processing task has been identified as a strong predictor of future alcohol problems (Schuckit et al., 2016). The frontal systems are involved in a host of cognitive and emotional processes including decision-making, affect and stress regulation, and problem solving, many of which have been linked with the risk for, or consequences of, AUDs (for a review see Moselhy et al., 2001). In addition, the anterior cingulate is important for adaptive decision making (Kennerley and Walton, 2011) and is often implicated in the processing of alcohol-related cues (Schacht et al., 2013). Thus, regionally specific alcohol-related CBF response in these frontal and anterior cingulate areas in a non-clinical sample of low LR young adults may represent another neurobiological marker of a more general cognitive/emotional AUD risk factor, such as the need for greater cognitive effort during task performance (Schuckit et al., 2016). However, the direction of effects in the present resting-state analysis do not seem to fit this model as high LR individuals tended to show enhanced alcohol versus placebo CBF response across the regions of study, suggesting greater resource utilization as compared to low LR individuals. Only within the low LR group do we see a statistically significant relationship between greater CBF response to alcohol and greater alcohol-related problems at follow-up. Further investigation of CBF in the context of other known risk markers, such as externalizing/internalizing profiles and impulsivity, may help to clarify how CBF function in these specific frontal and cingulate regions relate to alcohol-related behavioral outcomes.

Alternatively, given that the majority of studies demonstrating differential LR-related effects on BOLD markers of emotional and cognitive processing statistically controlled for CBF estimates in their analyses (Schuckit et al., 2012, Paulus et al., 2012, Schuckit et al., 2016), CBF response to alcohol may represent a neurophysiological process likely related to, yet independent from, these emotional and cognitive processes. For example, early rodent models have observed similar regionally specific effects on CBF following acute alcohol intoxication (Goldman et al., 1973), suggesting that tissue metabolism in these regions may increase at lower alcohol concentrations across species (Hoffman et al., 1986). Regionally specific increases in tissue metabolism may reflect the stimulating properties inherent to the consumption of alcohol (Newlin et al., 1982), yet greater research is needed to fully understand the physiological processes underlying these regional alcohol-induced effects. Additional CBF investigations incorporating measures of the stimulating responses to alcohol in humans (e.g., the brief Biphasic Alcohol Effects Scale [BAES]; Rueger and King, 2013) may be particularly insightful in this regard.

There are a number of additional factors that must be considered when interpreting these results. The sample size, while relatively large for neuroimaging studies, may be under powered to detect smaller effects of CBF and LR on alcohol outcomes and may have led to the non-significance of the other clusters investigated. Further, assessment of CBF during alternative time points along the blood alcohol curve may produce differential results. Replications of these results across the blood alcohol curve are required to determine if the pattern of cluster involvement is a stable finding. In addition, this sample is comprised of college students of higher socioeconomic backgrounds who were not experiencing clinical levels of AUD or other psychopathology at the time of CBF assessment. This limits our ability to generalize these results to more socioeconomically disadvantaged and clinically severe populations with AUD development prior to young adulthood. The results obtained in this sample of young adults should also be considered within the context of ongoing neurodevelopment of frontal systems (e.g., Simmonds et al., 2014) and may not generalize to older populations. Advantages of the study include a high follow-up rate (91% of the baseline sample) and thorough characterization of the LR profile.

In conclusion, this study adds to the growing body of literature supporting the link between CBF and important functional outcomes in AUDs (Durazzo et al., 2010, Goldstein et al., 2004). The results demonstrated a clinically relevant relationship between CBF and future alcohol problems, particularly in individuals with a low LR phenotype. Although the functional mechanisms underlying this relationship remain unclear, these initial results advance CBF as an important neurobiological marker for AUD risk, which may help to elucidate the relationship between the low LR phenotype and the development of later alcohol problems.

Acknowledgments

Support for this study was provided by NIH/NIAAA grants 1R21 AA021827 and 1R01 AA021162. KEC was supported by training grants from NIAAA (T32 AA013525) and NIMH (T32 MH018399). The authors have no conflicts of interest to declare.

References

  1. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). Washington, DC. [Google Scholar]
  2. Brown GG, Eyler Zorrilla LT, Georgy B, Kindermann SS, Wong EC, Buxton RB (2003) BOLD and perfusion response to finger-thumb apposition after acetazolamide administration: differential relationship to global perfusion. J Cereb Blood Flow Metab 23:829–37. [DOI] [PubMed] [Google Scholar]
  3. Bucholz KK, Cadoret R, Cloninger CR, Dinwiddie SH, Hesselbrock VM, Nurnberger JI Jr., Reich T, Schmidt I, Schuckit MA (1994) A new, semi-structured psychiatric interview for use in genetic linkage studies: a report on the reliability of the SSAGA. J Stud Alcohol 55:149–58. [DOI] [PubMed] [Google Scholar]
  4. Chung T, Martin CS (2009) Subjective stimulant and sedative effects of alcohol during early drinking experiences predict alcohol involvement in treated adolescents. J Stud Alcohol Drugs 70:660–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Claus ED, Ewing SW, Filbey FM, Sabbineni A, Hutchison KE (2011) Identifying neurobiological phenotypes associated with alcohol use disorder severity. Neuropsychopharmacology 36:2086–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cox RW (1996) AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput Biomed Res 29:162–73. [DOI] [PubMed] [Google Scholar]
  7. Cox RW, Chen G, Glen DR, Reynolds RC, Taylor PA (2017) FMRI Clustering in AFNI: False-Positive Rates Redux. Brain Connect 7:152–171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dick DM, Jones K, Saccone N, Hinrichs A, Wang JC, Goate A, Bierut L, Almasy L, Schuckit M, Hesselbrock V, Tischfield J, Foroud T, Edenberg H, Porjesz B, Begleiter H (2006) Endophenotypes successfully lead to gene identification: results from the collaborative study on the genetics of alcoholism. Behav Genet 36:112–26. [DOI] [PubMed] [Google Scholar]
  9. Durazzo TC, Gazdzinski S, Mon A, Meyerhoff DJ (2010) Cortical perfusion in alcohol-dependent individuals during short-term abstinence: relationships to resumption of hazardous drinking after treatment. Alcohol 44:201–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ehlers CL, Schuckit MA (1988) EEG response to ethanol in sons of alcoholics. Psychopharmacol Bull 24:434–7. [PubMed] [Google Scholar]
  11. Eklund A, Nichols TE, Knutsson H (2016) Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proc Natl Acad Sci 113:7900–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Eng MY, Schuckit MA, Smith TL (2005) The level of response to alcohol in daughters of alcoholics and controls. Drug Alcohol Depend 79:83–93. [DOI] [PubMed] [Google Scholar]
  13. Gmel G, Rehm J (2003) Harmful alcohol use. Alcohol Res Health 27:52–62. [PMC free article] [PubMed] [Google Scholar]
  14. Goldman H, Sapirstein LA, Murphy S, Moore J (1973) Alcohol and regional blood flow in brains of rats. Proc Soc Exp Biol Med 144:983–8. [DOI] [PubMed] [Google Scholar]
  15. Goldstein RZ, Leskovjan AC, Hoff AL, Hitzemann R, Bashan F, Khalsa SS, Wang GJ, Fowler JS, Volkow ND (2004) Severity of neuropsychological impairment in cocaine and alcohol addiction: association with metabolism in the prefrontal cortex. Neuropsychologia 42:1447–58. [DOI] [PubMed] [Google Scholar]
  16. Gundersen H, Wageningen HV, Grüner R (2013) Alcohol-Induced Changes in Cerebral Blood Flow and Cerebral Blood Volume in Social Drinkers. Alcohol Alcohol 48:160–165. [DOI] [PubMed] [Google Scholar]
  17. Hamdi E, Al-Suhaili A, Abou-Saleh MT, Amin Y, Prais V (2003) Cerebral blood flow in alcohol withdrawal: relation to severity of dependence and cognitive impairment. Acta Neuropsychiatr 15:55–62. [DOI] [PubMed] [Google Scholar]
  18. Hines LM, Ray L, Hutchison K, Tabakoff B (2005) Alcoholism: the dissection for endophenotypes. Dialogues Clin Neurosci 7:153–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hoffman WE, Miletich DJ, Albrecht RF (1986) Dose and time dependent cerebrovascular and metabolic effects of ethanol. Alcohol 3:23–6. [DOI] [PubMed] [Google Scholar]
  20. Hyman SE (1999) Introduction to the complex genetics of mental disorders. Biol Psychiatry 45:518–21. [DOI] [PubMed] [Google Scholar]
  21. Jueptner M, Weiller C (1995) Review: does measurement of regional cerebral blood flow reflect synaptic activity? Implications for PET and fMRI. Neuroimage 2:148–56. [DOI] [PubMed] [Google Scholar]
  22. Kennerley SW, Walton ME (2011) Decision making and reward in frontal cortex: complementary evidence from neurophysiological and neuropsychological studies. Behav Neurosci 125:297–317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Khalili-Mahani N, Van Osch MJP, Baerends E, Soeter RP, De Kam M, Zoethout RWM, Dahan A, Van Buchem MA, Van Gerven JMA, Rombouts S (2011) Pseudocontinuous arterial spin labeling reveals dissociable effects of morphine and alcohol on regional cerebral blood flow. J Cereb Blood Flow Metab 31:1321–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. King AC, Mcnamara PJ, Hasin DS, Cao D (2014) Alcohol challenge responses predict future alcohol use disorder symptoms: a 6-year prospective study. Biol Psychiatry 75:798–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Marxen M, Gan G, Schwarz D, Mennigen E, Pilhatsch M, Zimmermann US, Guenther M, Smolka MN (2014) Acute effects of alcohol on brain perfusion monitored with arterial spin labeling magnetic resonance imaging in young adults. J Cereb Blood Flow Metab 34:472–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Moselhy HF, Georgiou G, Kahn A (2001) Frontal lobe changes in alcoholism: a review of the literature. Alcohol Alcohol 36:357–68. [DOI] [PubMed] [Google Scholar]
  27. Murray DE, Durazzo TC, Mon A, Schmidt TP, Meyerhoff DJ (2015) Brain perfusion in polysubstance users: relationship to substance and tobacco use, cognition, and self-regulation. Drug Alcohol Depend 150:120–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Newlin DB, Golden CJ, Quaife M, Graber B (1982) Effect of alcohol ingestion on regional cerebral blood flow. Int J Neurosci 17:145–50. [DOI] [PubMed] [Google Scholar]
  29. Paulus MP, Schuckit MA, Tapert SF, Tolentino NJ, Matthews SC, Smith TL, Trim RS, Hall SA, Simmons AN (2012) High versus low level of response to alcohol: evidence of differential reactivity to emotional stimuli. Biol Psychiatry 72:848–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Piano MR, Mazzuco A, Kang M, Phillips SA (2017) Cardiovascular Consequences of Binge Drinking: An Integrative Review with Implications for Advocacy, Policy, and Research. Alcohol Clin Exp Res 41:487–496. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P (1999) SENSE: sensitivity encoding for fast MRI. Magn Reson Med 42:952–62. [PubMed] [Google Scholar]
  32. Raichle ME, Grubb RL Jr., Gado MH, Eichling JO, Ter-Pogossian MM (1976) Correlation between regional cerebral blood flow and oxidative metabolism. In vivo studies in man. Arch Neurol 33:523–6. [DOI] [PubMed] [Google Scholar]
  33. Ray LA, Mackillop J, Monti PM (2010) Subjective Responses to Alcohol Consumption as Endophenotypes: Advancing Behavioral Genetics in Etiological and Treatment Models of Alcoholism. Subst Use Misuse 45:1742–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Restom K, Bangen KJ, Bondi MW, Perthen JE, Liu TT (2007) Cerebral blood flow and BOLD responses to a memory encoding task: a comparison between healthy young and elderly adults. Neuroimage 37:430–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rueger SY, King AC (2013) Validation of the brief Biphasic Alcohol Effects Scale (B-BAES). Alcohol Clin Exp Res 37:470–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sacks JJ, Gonzales KR, Bouchery EE, Tomedi LE, Brewer RD (2015) 2010 National and State Costs of Excessive Alcohol Consumption. Am J Prev Med 49:e73–e79. [DOI] [PubMed] [Google Scholar]
  37. Savage JE, Neale Z, Cho SB, Hancock L, Kalmijn JA, Smith TL, Schuckit MA, Donovan KK, Dick DM (2015) Level of response to alcohol as a factor for targeted prevention in college students. Alcohol Clin Exp Res 39:2215–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Schacht JP, Anton RF, Myrick H (2013) Functional neuroimaging studies of alcohol cue reactivity: a quantitative meta-analysis and systematic review. Addict Biol 18:121–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Schuckit M, Smith T, Paulus M, Tapert S, Simmons A, Tolentino N, Shafir A (2016) The Ability of Functional Magnetic Resonance Imaging to Predict Heavy Drinking and Alcohol Problems 5 Years Later. Alcohol Clin Exp Res 40:206–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Schuckit MA (2009) An overview of genetic influences in alcoholism. J Subst Abuse Treat 36:S5–14. [PubMed] [Google Scholar]
  41. Schuckit MA, Gold E, Risch C (1987) Serum prolactin levels in sons of alcoholics and control subjects. Am J Psychiatry 144:854–9. [DOI] [PubMed] [Google Scholar]
  42. Schuckit MA, Gold EO, Croot K, Finn P, Polich J (1988a) P300 latency after ethanol ingestion in sons of alcoholics and in controls. Biol Psychiatry 24:310–5. [DOI] [PubMed] [Google Scholar]
  43. Schuckit MA, Risch SC, Gold EO (1988b) Alcohol consumption, ACTH level, and family history of alcoholism. Am J Psychiatry 145:1391–5. [DOI] [PubMed] [Google Scholar]
  44. Schuckit MA, Smith TL, Danko GP, Pierson J, Hesselbrock V, Bucholz KK, Kramer J, Kuperman S, Dietiker C, Brandon R, Chan G (2007) The ability of the Self-Rating of the Effects of Alcohol (SRE) Scale to predict alcohol-related outcomes five years later. J Stud Alcohol Drugs 68:371–8. [DOI] [PubMed] [Google Scholar]
  45. Schuckit MA, Smith TL, Tipp JE (1997) The Self-Rating of the Effects of alcohol (SRE) form as a retrospective measure of the risk for alcoholism. Addiction 92:979–88. [PubMed] [Google Scholar]
  46. Schuckit MA, Smith TL, Trim RS, Heron J, Horwood J, Davis J, Hibbeln J (2008) The self-rating of the effects of alcohol questionnaire as a predictor of alcohol-related outcomes in 12-year-old subjects. Alcohol Alcohol 43:641–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Schuckit MA, Tapert S, Matthews SC, Paulus MP, Tolentino NJ, Smith TL, Trim RS, Hall S, Simmons A (2012) fMRI differences between subjects with low and high responses to alcohol during a stop signal task. Alcohol Clin Exp Res 36:130–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Shieh G (2011) Clarifying the role of mean centring in multicollinearity of interaction effects. Br J Math Stat Psychol 64:462–77. [DOI] [PubMed] [Google Scholar]
  49. Simmonds D, Hallquist MN, Asato M, Luna B (2014) Developmental Stages and Sex Differences of White Matter and Behavioral Development through Adolescence: A Longitudinal Diffusion Tensor Imaging (DTI) Study. Neuroimage 92:356–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE, Niazy RK, Saunders J, Vickers J, Zhang Y, De Stefano N, Brady JM, Matthews PM (2004) Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 23 Suppl 1:S208–19. [DOI] [PubMed] [Google Scholar]
  51. Squeglia LM, Jacobus J, Tapert SF (2014) The effect of alcohol use on human adolescent brain structures and systems. Handb Clin Neurol 125:501–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Strang NM, Claus ED, Ramchandani VA, Graff-Guerrero A, Boileau I, Hendershot CS (2015) Dose-dependent effects of intravenous alcohol administration on cerebral blood flow in young adults. Psychopharmacology 232:733–44. [DOI] [PubMed] [Google Scholar]
  53. Substance Abuse and Mental Health Services Administration (2017). Key substance use and mental health indicators in the United States: Results from the 2016 National Survey on Drug Use and Health. Rockville, MD. [Google Scholar]
  54. Tapert SF, Pulido C, Paulus MP, Schuckit MA, Burke C (2004) Level of response to alcohol and brain response during visual working memory. J Stud Alcohol 65:692–700. [DOI] [PubMed] [Google Scholar]
  55. Tiihonen J, Kuikka J, Hakola P, Paanila J, Airaksinen J, Eronen M, Hallikainen T (1994) Acute ethanol-induced changes in cerebral blood flow. Am J Psychiatry 151:1505–8. [DOI] [PubMed] [Google Scholar]
  56. Tolentino NJ, Wierenga CE, Hall S, Tapert SF, Paulus MP, Liu TT, Smith TL, Schuckit MA (2011) Alcohol effects on cerebral blood flow in subjects with low and high responses to alcohol. Alcohol Clin Exp Res 35:1034–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Trim RS, Simmons AN, Tolentino NJ, Hall SA, Matthews SC, Robinson SK, Smith TL, Padula CB, Paulus MP, Tapert SF, Schuckit MA (2010) Acute ethanol effects on brain activation in low- and high-level responders to alcohol. Alcohol Clin Exp Res 34:1162–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Weiger M, Pruessmann KP, Osterbauer R, Bornert P, Boesiger P, Jezzard P (2002) Sensitivity-encoded single-shot spiral imaging for reduced susceptibility artifacts in BOLD fMRI. Magn Reson Med 48:860–6. [DOI] [PubMed] [Google Scholar]
  59. Wong EC (2005) Quantifying CBF with pulsed ASL: technical and pulse sequence factors. J Magn Reson Imaging 22:727–31. [DOI] [PubMed] [Google Scholar]
  60. Zhang A, Fisher AJ, Bailey JO, Kass AE, Wilfley DE, Taylor CB (2015) The self-rating of the effects of alcohol questionnaire predicts heavy episodic drinking in a high-risk eating disorder population. Int J Eat Disord 48:333–6. [DOI] [PubMed] [Google Scholar]
  61. Zheng H, Kong L, Chen L, Zhang H, Zheng W (2015) Acute effects of alcohol on the human brain: a resting-state FMRI study. Biomed Res Int 2015:947529. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES