Alterations of brain activity in patients with alcohol use disorder: a resting-state fMRI study

Alcohol use disorder (AUD), one of the most prevalent psychiatric disorders in the world, is defined by compulsive drinking behavior and continuous, excessive, uncontrolled alcohol consumption [1]. AUD is closely related to the problems of public health, social, criminal, and mental health [2]. Continued excessive drinking can have an adverse effect on several organs and is also directly related to a number of malignancies [3].

A number of novel biomarkers for the diagnosis of AUD have been identified through brain imaging techniques in recent years. Voxel-based morphometry (VBM) studies revealed that alcohol consumption was significantly associated with lower total and local gray matter volume [4, 5]. Reduced grey matter volume is associated with the duration time of alcohol dependence and lifetime alcohol consumption [6]. Whole-brain voxel-wise analyses of diffusion tensor imaging (DTI) found that AUD patients had extensive damage to white matter (WM) structures [7]. AUD not only causes structural damage but also functional changes in brain regions. In a proton Magnetic Resonance Spectroscopy (MRS) study, alcohol intake was negatively correlated with Choline/Creatine ratio in the left prefrontal cortex, perhaps indicating the severity of alcohol abuse [8]. The left parietal region and the prefrontal area were also less perfused in AUD patients [9]. Prior research by our team revealed that the efficiency of the brain network in the AUD group was significantly lower [10].

The resting-state functional magnetic resonance imaging (rs-fMRI) technique is a non-invasive method that is widely used to quantify the neural network structure and functional brain properties of AUD patients [11]. Binarized and weighted degree centrality (DC) can reflect the density and strength of whole-brain functional connectivity (FC), which has the advantages of avoiding the influence of subjective selection by not defining the seed points, and can indirectly reflect the importance and changes of brain network nodes [12]. Many diseases have been studied using the voxel-wise DC approach, but the use of binarized and weighted DC methods in the study of AUD patients has been limited. This study used rs-fMRI to track DC and FC changes in brain activity in AUD patients and correlate these changes with clinical and neuropsychological information.

In this study, we investigated alterations in binarized, weighted degree centrality and functional connectivity in AUD patients. Firstly, we used binarized and weighted DC methods to analyze the differences of the nodes in brain network between the two groups. The AUD group had significantly different binarized and weighted DC values in the left PreCG and the left IPL compared to HC group. Then, changes in the seed-based FC were used to further validate the results of DC. FC analyses indicated that AUD groups showed a decreased left PreCG/IPL-based FC in the right ITG, MTG, IOG, SPG, cerebellum_Crus 1, the left SFGdor, PCUN, LING and bilateral IPL. In this study, PreCG and IPL might be closely related to the occurrence of AUD. Through the statistical examination of relevant data, we found a statistically significant association between altered FC and clinical symptoms. Particularly, the FC values of the right MTG - right PreCG as well as the right IPL - left IPL were linked with MAST and ADS, respectively. This suggests that alternations of FC might be associated with AUD.

The AUD group had significantly different DC and FC values in the PreCG and the IPL. The PreCG region is known as the primary motor center of the brain [18], involved in the integration of information related to the sensory, motor, attention, and reward circuits [19‐22]. Our findings of reduced DC and FC in the left PreCG region of AUD patients further enriched the evidence that chronic alcohol consumption might cause functional changes in the PreCG, which would affect sensory, motor and attention functions as well as brain reward mechanisms. And IPL is involved in attention allocation and assessment of decisions [23, 24]. Results from the present study suggest that patients with AUD have lower FC in a network involved in attention, which may account for the inability of AUD patients to make correct decisions and assessments. The cerebellum damage is also a risk with alcohol consumption. The most common central nervous system complication of AUD is cerebellar degeneration, which primarily manifests as cerebellar ataxia in approximately 10–25% of AUD [25]. Previous studies have reported extensive atrophy of the cerebellar gray matter in AUD patients [26]. AUD is associated with brain injury, particularly the frontal-parietal cerebellar gyrus, a circuit abnormality associated with changes in language, spatial working memory, executive function, gait and balance [27]. Alcohol-induced brain differences have also been observed in the right IPL with significantly higher amplitude of low frequency fluctuation (ALFF) values, conversely the ALFF values were lower in the posterior lobe of cerebellum and the left PCUN, suggesting disruption of the prefrontal-parietal-cerebellar circuit in individuals with AUD [28]. In this study, we found decreased functional connectivity in the left PreCG - left SFGdor, left IPL - right cerebellum_crus1/SPG, and left IPL - left PCUN in patients with AUD. This may indicate abnormal damage to the PreCG/IPL region, which affects the prefrontal-parietal-cerebellar circuit.

Reduced connectivity in the PCUN, LING, and IPL characterized the resting state functional network connectivity driven by drinking. These results are consistent with previous hypoconnectivity observations in visual and sensorimotor areas [29, 30]. Furthermore, as previously found, we identified the PCUN as a region associated with AUD [31]. The PCUN is thought to be engaged in visuospatial imagery, collection and evaluation of information, and self-processing operations [32‐34]. Disturbed regional brain activity or functional connectivity in the PCUN was found in subjects with AUD and after heavy drinking [30, 35, 36]. Abnormal activation of the PCUN in subjects with AUD is thought to be associated with craving and visual memory processing [37]. The functional brain activity decrease in the PCUN may be interpreted as functional impairment caused by AUD. The significant FC differences were also observed in the right IOG and the left LING. These regions are involved in visual information processing and visual processing impairment is a common symptom in patients with AUD [38]. Previous research conducted by our team also revealed that impaired node efficiency was observed in visual cortex regions [10]. The visual dysfunction is prevalent in drinkers, however, the visual dysfunction associated to drinking is located around the LING, in contrast, the FC differences driven by both smoking and drinking are more frequently located about IOG [30]. Additionally, there is a correlation between activity in the visual cortex and vividness of mental imagery, suggesting that fMRI can detect individual differences in the vividness of mental imagery [39].

The AUD group exhibited a decrease in FC within SFGdor and SPG regions. The changes of these regions are consistent with previous studies and earlier work of our team [10, 40]. However, The ReHo study showed that SFGdor, PreCG, ITG and MTG are among the regions with stronger resting-state connections in individuals suffering from AUD [41]. Relapse after abstinence and reliance on alcohol can also be connected to aberrant frontal lobe neuronal activity in AUD patients, which may be related to failed abstinence behaviors in AUD patients [42, 43].

The regulation of the temporal lobe involves both vision and auditory senses [44]. The MTG is responsible for language-auditory processing, and the ITG is associated with the representation and detection of complex object features [45]. Long-term alcohol dependence is associated with compulsive drug-seeking behavior, which may stimulate the verbal, visual and auditory brain regions. Research by Bach et al. [46] demonstrated that patients with AUD exhibited a decrease in gray matter volume in bilateral MTG and ITG, which are primarily responsible for processing visual information and may be influenced by alcohol dehydrogenase genes. The interference of the brain reward network and ventral visual system with alcohol can result in affective changes and a decline in visual function [47]. The MTG, as part of the emotional circuit, plays a role in emotional responses. In an fMRI study on brain reactivity in response to emotional stimuli, the AUD group showed reduced responses to fear and disgust in the right amygdala, right MTG, pons, and left superior frontal gyrus, consistent with previous findings [48]. According to neurophysiological studies, the inferior temporal lobe played a role in the recognition of objects and the ventral stream of vision [49]. It is possible that the reduced connectivity in the right ITG correspondence correlates with impaired visual function, selectively affecting object discrimination and face recognition. In this study, weakened functional connectivity was found in the left PreCG - right ITG/MTG and left IPL - right ITG of AUD patients, potentially linked to impaired information communication and integration between corresponding brain regions, or abnormalities in the emotional circuit system. These findings may contribute to emotional instability, decreased attention, and changes in visual and auditory perception.

The dysconnectivity of the left PreCG - right MTG and the right IPL - left IPL might be possibly connected to clinical outcomes in AUD patients, as evidenced by the FC value of the left PreCG - right MTG being adversely correlated with MAST scores. Interestingly, we also discovered a positive correlation between the ADS scores and the FC value of the right IPL - left IPL, meaning that the AUD patients with lower FC values has a lighter dependence on alcohol. A previous study [11] showed that when patients with AUD had more severe cognitive impairment, as a compensation, they had less alcohol dependence. In conjunction with the current study, we can speculate that AUD patients with lower right IPL - left IPL FC values are less dependent on alcohol, which may be related to more severe cognitive impairment. There may be compensatory mechanisms between symptom severity and cognitive impairment, suggesting a difference in pathological pathways. These findings assisted future AUD studies by providing insight into possible mechanisms. Although we did not take the ADS score of HC group, HC has the higher FC values of right IPL-left IPL, which means the correlation between ADS and FC values of right IPL-left IPL cannot fit in HC group. These alterations in the mechanism between HC and AUD indicate an unseen functional role played by the right IPL-left IPL, which needs further explorations in future studies.

There were some limitations to be noted. Firstly, the sample size is relatively small, although rs-fMRI studies with more than 16 subjects per group are considered acceptable [50]. Secondly, this study is cross-sectional, which means we cannot directly link anomalous connections to AUD. Further longitudinal studies would be beneficial in addressing this issue. Thirdly, in our study, the cutoff for DC analysis was only set at r > 0.25 [36]. More thresholds should be used for comparison in future studies to reduce the effect of methodological choices on results. Fourthly, multiple analyses of psychometric-neural correlation data were not performed due to sample size, the purpose and methodology of this study.

In summary, this study aimed to investigate the brain network of AUD patients using binarized DC, weighted DC and FC. The results showed extensive similarities in the spatial distribution of binarized and weighted DC values between the AUD and HC groups. Alterations in DC values in the brain network of AUD patients were found mainly in the PreCG and IPL. Besides, the FC within the in the right ITG, MTG, IOG, SPG, cerebellum_Crus 1, the left SFGdor, PCUN, LING and bilateral IPL were damaged in patients with AUD as well. Furthermore, the correlations between behavioral and node connectivity suggested that multiple inter-network connectivity dysfunctions played a compensatory role in AUD. DC and FC methods are valuable in detecting biomarkers / brain activity features specific for the AUD. These findings contribute to further understanding toward the brain network activities in patients with AUD.