Betel quid dependence is associated with functional connectivity changes of the anterior cingulate cortex: a resting-state fMRI study

In Hainan province of China, Betel quid (BQ) is the combination of fresh Areca nut (AN) and slaked lime (i.e. aqueous calcium hydroxide paste) wrapped in a betel leaf without tobacco and other ingredients. BQ is chewed by approximately 600 million people globally, with its use concentrated in South Asia, Southeast Asia, and Pacific islands [1]. BQ chewing is an important behavior from a public health perspective because it is associated with a variety of health issues, most notably oral cancer and precancerous conditions such as leukoplakia and oral submucous fibrosis [2]. Consequently, BQ has been categorized into a Group 1 carcinogen by the International Agency for Research on Cancer [1, 3]. BQ ranks the fourth most frequently consumed psychoactive substance around the globe, following only alcohol, nicotine, and caffeine in prevalence [4]. BQ use is related to a dependency syndrome, which is characterized by enhanced concentration, mild euphoria, relaxation, postprandial satisfaction and a withdrawal syndrome involved with sleeplessness, unstable mood, irritability and anxiety, and its severity can be similar to that of amphetamine [5]. However, the mechanism by which the BQ dependence (BQD) behaviour influences cerebral function through working on the specific cerebral areas has not yet been clearly illuminated. Our previous voxel based morphometry (VBM) study of BQD indicated that BQD individuals have gray matter volumes decreased in the right rostral anterior cingulate cortex (ACC), which also indicated a negative correlation with the duration of BQD [6]. These findings might be helpful to investigate potential structural substrates of the BQD.

The use of chronic addictive drug is usually associated with abnormal functional organization in the user’s brain, which results in executive dysfunction [7] and disinhibition [8] to the drug and drug-related cues and makes sure their compulsive patterns of drug-seeking actions [9]. The connection of spontaneous fluctuations of blood oxygen level-dependent signals in diverse areas of the “resting” brain assesses the resting-state functional connectivity (rsFC) which is considered as a measure of its functional organization [10]. Studies have outlined a number of resting-state networks corresponding to key brain functional organizations including vision, movement, language, audition, executive function, episodic memory, and salience detection [10]. It is thought that the default mode network (DMN) [11], the most famous resting-state network consisting of brain regions deactivated during external-oriented tasks, is involved in the maintenance of the baseline brain activities related to cognitions of self-awareness, episodic memory and interactive modulation between the interior mind activities and exterior tasks [12]. The sensorimotor network (SMN) [13], which is significant for the performance of voluntary movements, functionally connects regions within the primary motor cortex (M1) and the supplementary motor area (SMA) [14, 15]. The affective network (AN) comprises a corticolimbic circuit responsible for negative emotional arousal or regulation [16–18] and participates in autonomic and visceral functions [19, 20]. Respectively, the cognitive network (CN) and the visual network (VN) cope with corresponding cognition and vision functions [21].

Since it is more readily applicable than functional activation MRI in clinical environment, several groups have started to analyze the rsFC in various neuropsychiatric diseases such as Alzheimer’s disease and depression [22, 23]. According to models of addiction, the major brain regions underlying addiction constitute a network of at least four interdependent and overlapping circuits [24]: (i) motivation and/or drive and salience evaluation, located in the orbital frontal cortex; (ii) cognitive control, located in the dorsal anterior cingulate cortex and prefrontal cortex; (iii) memory and learning, including the hippocampus and amygdala; and (iv) reward, involving the ventral pallidum and nucleus accumbens. In addition, the ACC is an essential part of the frontal-subcortical circuit, and it plays a modulatory role in cognition, emotion and reward expectations [16, 25–27], and is also likely to affect the reactivity of the above circuits and parts of this network [16, 28, 29]. Convergent evidence from lesion and neuroimaging studies consistently indicate that drug dependence is closely associated with structural [30], functional [31], and metabolic [32] abnormalities of the ACC. Recent molecular and neuroimaging studies have further theorized that altered baseline or resting-state activity in the ACC is associated with drug dependence [33–35].

To date, the rsFC within the key regions of BQD has not been extensively studied. Therefore, in this study we detected whether there was alteration of functional connectivity related to ACC on samples with BQD for investigating the mechanisms underlying BQD.

38 BQD individuals and 36 healthy control subjects were recruited for this study. 3 BQD individuals with angio cavernoma, arachnoid cyst and lacunar infarction respectively were excluded for lesions of the brain. Two BQD individuals and four healthy control subjects were excluded because of excessive head movement.

Within-group comparison of the ACC resting-state functional network

Correlation maps were produced by extracting the BOLD time course from bilateral ACC then computing the correlation coefficient between that time course and the time course from all other brain voxels. Positive z-score values are significantly correlated or positive with the seed region, whereas negative z-score values are significantly anti correlated or negative [51].

The results from the one-sample t test showed the spatial distribution of functional connectivity patterns of bilateral ACC with the whole brain for all subjects were similar, although potentially interesting variations can be seen (Fig. 1). The functional connectivity patterns of ACC were also very similar in the right and left hemispheres. In all subjects, the positive functional connectivity with ACC (orange) mainly involved (i) the DMN, including medial prefrontal cortex, posterior cingulate cortex (PCC)/precuneus, the middle and inferior temporal gyri (MTG/ITG), and superior frontal gyrus; (ii) the AN covered the orbitofrontal cortex, medial prefrontal cortex, and temporal pole, the thalamus, insula, putamen and caudate (P < 0.05, FDR corrected) (Fig. 1). On the other hand, the negative functional connectivity with the seed regions (blue) mainly involved (i) the SMN, which included the precentral and postcentral gyri; (ii) the CN, which included the dorsolateral prefrontal cortex (dlPFC), ventrolateral prefrontal cortex (vlPFC) and dorsolateral parietal cortex; and (iii) the VN, which included the cuneus lobe, and lingual and fusiform gyri (P < 0.05, FDR corrected) (Fig. 1).

Inter-group comparisons of the ACC resting-state functional network

Functional connectivity differences between HCs and BQD from two-sample t tests, are displayed in Table 2 and Fig. 2 (Alpha Sim corrected, P < 0.01).The individuals with BQD indicated significant increases in the FC (orange) between bilateral ACC and pons, between bilateral ACC and caudate, between bilateral ACC and bilateral thalamus, between bilateral ACC and midbrain, between bilateral ACC and cerebellum (Fig. 2). By contrast, bilateral ACC exhibited decreased functional connectivity (blue) to the medial prefrontal cortex (mPFC), the left precuneus, and the left para Hippocampal/hypothalamus (Fig. 2). Critically, the right ACC maps are to a large extent sign-inverted versions of the left ACC maps (Fig. 2).

Brain region	BA	Voxels	Maximal t value	MNI coordinates
				X	Y	Z
Left ACC
Pons		61	5.23	−6	−27	−33
L. caudate		38	4.26	−12	3	12
L. thalamus		53	3.44	−6	−18	12
R. thalamus		51	3.37	9	−21	9
Midbrain		43	2.98	−6	−12	−12
Cerebellum posterior lobe		27	3.10	−42	−57	−36
vmPFC, ACC	47	424	−4.5	15	30	−9
L. precuneus	7	32	−3.68	−15	−57	57
L. parahippocampal/hypothalamus		33	−3.36	−12	0	−18
Right ACC
Pons		114	5.87	−6	−24	−36
L. caudate		43	4.90	−12	3	12
R. caudate		41	3.45	15	15	9
L. thalamus		117	3.95	−6	−18	12
R. thalamus		64	3.91	18	−9	6
Midbrain		88	4.18	6	−9	−9
L. cerebellum posterior lobe		49	4.07	−36	−66	−33
Cerebellum anterior lobe		62	3.87	6	−60	−30
R. cerebellum posterior lobe		53	3.73	30	−66	−36
mPFC, ACC	32.47	461	−4.47	18	27	−27
L. precuneus	7	24	−3.80	−15	−57	60
hypothalamus/parahippocampal_L		24	−3.35	3	0	−15

A corrected threshold of P < 0.01 determined by Monte Carlo simulation was taken as meaning that there was a significant difference between groups

L left, R right, mPFC medial prefrontal cortex, ACC anterior cingulate cortex, BA Brodmann’s area, MNI montreal neurological institute, x, y, z coordinates of primary peak locations in the MNI space

Correlation analysis results

In BQD individuals, the functional connectivity of right ACC to left thalamus and left ACC to the pons were significantly and specifically positively correlated with BQDS (r = 0.459; P = 0.007; r = 0.359; P = 0.040) (Fig. 3). In addition, the functional connectivity of right ACC to the left precuneus were negatively associated with durations (r = −0.403, P = 0.020) (Fig. 3). The other increased or decreased functional connectivity to the ACC were separated from BQDS and durations.

In the current rsFC study, we investigated the potential alterations of functional connectivity related to ACC on BQD and the correlation of functional connectivity changes with the severity and duration of BQD. Interestingly, we found the number of male individuals were higher than that of female individuals in BQD group. The functional connectivity results show that the aberrant functional connectivity during resting state between ACC and regions of reward and cognitive control network, which is associated with the neural mechanism of BQD. It has been reported that the cause of addiction is not onefold—including genetic—but that genes do occupy a vital position “by influencing processes in the brain and body [52–54] that interact with one another and with an individual’s life experiences to produce susceptibility or protection “[55], so the sex difference and structural plasticity in vulnerability to addiction may be implicated in those genetic backgrounds. These, of course, need further researches.

In this study, the ACC functional connectivity patterns mapped in HCs, and BQD individuals were in line with those reported in the previous study [56]. The ACC functional connectivity positively correlated with the DMN, AN, and brain regions that process different aspects of emotionally salient stimuli, the thalamus, caudate, and putamen. The ACC functional connectivity negatively correlated with the SMN, CN, VN. All subjects had the similar functional connectivity patterns of the ACC. However, BQD individuals had intensity distribution difference in ACC functional connectivity patterns. This suggests that our findings on the ACC functional connectivity are reliable.

The results of rsFC by two-sample t tests, measured in whole brain with seeds in the bilateral ACC, indicate that dysfunctional integrations take place in the brains of resting-state BQD individuals. Compared to the HC group, the BQD group demonstrated greater connectivity from ACC to the regions of the reward network (midbrain regions, pons, caudate, thalamus). Specifically, the BQDS were positively correlated with increased functional connectivity of right ACC to left thalamus and left ACC to pons. The reward circuit, a core component for the development and monitoring of motivated behaviors, is now regarded to be embedded within the cortico-basal ganglia network. The role of the ACC has led to its inclusion in many major theories of addiction, where it is believed to form part of an inhibitory system that exercises control over reward-related behavior [57, 58]. Recent research has indicated that the midbrain and striatal areas which are involved in the reward remain more extensive than previous estimates. An increasing number of fMRI researches have started to study reward-processing midbrain/pons activity in individuals [59]. Midbrain dopamine (DA) neurons play a central role in a wide range of behaviors, from attention and motivation to motor control and reinforcement. Increased midbrain activation during anticipation of pleasant tastes [60, 61], anticipation of monetary gains [62], and during exposure to visual stimuli that evoke romantic love [63] have been reported by task-based fMRI researchers. The dorsal striatal circuitry plays a pivotal role during the development of habitual compulsive drug dependence [64]. Brain imaging research indicated that dopamine increases in the dorsal striatum (caudate and putamen) were induced by drug-associated cues, which was an effect that related to self-reports of craving [65, 66]. Previous study has not covered the thalamus extensively in terms of addiction. Nevertheless, this area has been increasingly implicated during addiction due to its integrative function in the regulation of arousal and attentional modulation. For instance, dopamine neurotransmission in the thalamus was increased with intravenous administration of a stimulant drug in cocaine users rather than in controls, which was an effect related to craving [67]. Both PET and fMRI results demonstrate that thalamic activation can be increased by primary and secondary rewards (vs. non rewards) [68]. The ACC-thalamus functional connectivity were increased compared with HC; together with the finding that the right ACC-left thalamus positively correlated with BQDS, indicated the thalamus can serve as a specific cerebral area during the BQD development. Based on the previous studies and our results above, we argue that the functional connectivity in ACC to the reward network may be correlated with cognitive management and behavioral dysfunctions directed by goal in BQ addiction at least in some degree [69, 70]. Although BQ is a weakly addictive substance, our outcomes suggested that which shares similar functional organization to other addictions related to substance, indicating a more general pathomechanism for diseases correlated with increased BQD.

Interestingly, we found increased functional connectivity between ACC with cerebellum in BDQ group. It was observed in Previous PET and fMRI studies that drug-conditioned cue elicited the increased metabolism and activation of cerebellum [71, 72]. Glucose metabolism was greatly increased in the cerebellum when addicts performed reward expectation tasks [73], which suggested that the cerebellum is also included in drug-conditioned memories in addicts. And the cerebellum has been noted and described to play a compensatory role in inhibitory control [74] and decision-making behavior in addicts [75]. Based on the previous studies and our results above, we argue that the altered functional connectivity in ACC with cerebellum might be a reflection of the neuro adaptation and reorganization of cerebral functional network caused by BQD. These, of course, need further researches.

In addition, lower connectivity was observed in the BQD group in regions from ACC to the regions of the DMN [76], including mPFC and precuneus, which is part of the posterior parietal cortex. DMN might be the most important component of brain network at resting state [11] while it is a set of brain regions (e.g. ACC, mPFC and precuneus) that are identified as deactivated during demanding cognitive tasks and anti-correlated with the fronto-parietal regions [77–80]. The function of DMN is largely connected with self-relevant, internally directed information processing [79]. Studies of heroin and cocaine dependent individuals have found that hypo activation in the rostral ACC and mPFC is associated with deficits in response inhibition and impulse control [81, 82]. Li indicated that adequate inhibitory control and low craving levels may be caused by an increase of processing in the PFC-ACC, and in states with hypoactivity of the PFC-ACC circuits increased activity of the stress and reward centers, which results in a greater susceptibility to drug use and relapse [82]. Therefore, it can be concluded that disrupting monitoring function in resting-state ACC/mPFC may lead to aberrant behavior in BQD individuals. The functional connectivity of right ACC-left precuneus has a negative correlation with the durations supported the role of precuneus executive control dysfunction in BQD individuals, the BQD may need greater functional connectivity of precuneus to keep executive and inhibitory control balance, but it was further decreased with the durations of BQD. Overall, this study could exhibit the impairment of DMN in BQD individuals from a perspective of functional connectivity.

It was recognized that limbic regions connected with the hypothalamus could also influence addictive behaviors [83]. The lateral hypothalamus was intimately involved in reward, and the hypothalamic peptides, including galanin, enkephalin, orexin and melanin-concentrating hormone, could also stimulate alcohol intake [84–86]. But our outcomes of functional connectivity in regions from ACC to the left para Hippocampal/hypothalamus were decreased, so we argue that might be a reflection of abnormal inhibitory function related to ACC to mesocorticolimbic regions and hypothalamus caused by BQD.

The current study has several limitations. First, the abnormal functional connectivity with ACC in the BQD can only be observed as a cross-sectional study. However, we cannot draw direct causal inferences in terms of the relations between the brain disconnections and BQD. Hence, it would be beneficial to make further longitudinal studies with fMRI experiments in order to set up the cause—effect relations. Second, there was no diagnostic criterion for BQD-related executive dysfunction or disinhibition, and our understanding of the results was limited by such lack of objective and specific neurocognitive assessment. Should such clinical assessment become available, would allow researchers to explore BQD individuals’ brain patterns. More sophisticated clinical assessment (such as Digit Symbol Test, Trail Making Test and the Stroop Color-Word Test to measure cognitive function, Barratt Impulsiveness Scale to compared self-report impulsivity), may provide more accurate interpretation of relative functional connectivity characteristics. Third, given the research indicating that ACC abnormalities play a role in addiction more generally, it is likely that reduced right ACC spontaneous brain activity would also confer risk for other aspects of problematic substance use. Our findings need to be further verified with other drug dependence. Future explorations are still needed to confirm the conclusions.

Generally, the circuits make interactions and operate simultaneously. However, we found there were significant differences in them between BQD and HC, which is possibly due to the deficiency of reciprocity in these circuits and dysregulation of the integrative systems of rsFC in BQD individuals, and therefore results in maladaptive behaviors. To our understanding, this research is the first rs-fMRI investigation demonstrating an association of BQD behavior with the ACC functional connectivity. Despite those limitations, the abnormal cerebral functional connectivity can also contribute to revealing the neuropathologic mechanisms of BQD. Individuals with BQD develop increased ACC functional connectivity to the areas of the reward network and decreased to the DMN, especially increased functional connectivity in the right ACC- left thalamus and left ACC-pons, may, therefore, represent a biomarker of BQD individuals. Additionally, disturbances in functional connectivity during resting state from ACC to the whole brain revealed by fMRI may have a key function in providing insights into the neurological pathophysiology underlying BQD-associated executive dysfunction and disinhibition. Such results may contribute to our better understanding of the mechanisms of BQD. And further researches and analyses will be carried out for the purpose of detecting the reciprocal functions of each circuit which is affected by BQD.