Quantitative Electroencephalogram in Persons with Alcohol Dependence Syndrome: A Pre- and Post-Treatment Findings

Abstract

Background

Assessment and treatment monitoring in alcohol dependence syndrome often rely on subjective measures, particularly in resource-limited settings. Quantitative electroencephalogram (qEEG) provides an objective alternative, though its role in alcohol use and abstinence remains underexplored in the Indian context.

Aim

To study changes in the quantitative electroencephalogram in persons with alcohol dependence syndrome undergoing treatment.

Methods

Patients diagnosed with alcohol dependence syndrome as per ICD-11 were recruited. At baseline, the Severity of Alcohol Dependence Questionnaire (SADQ) (mean 22.60 ± 4.81) and Clinical Institute Withdrawal Assessment for Alcohol–Revised (CIWA-Ar) (mean 10.98 ± 2.45) were administered to assess the severity of dependence and withdrawal symptoms. qEEG was recorded at baseline, following detoxification and at 12 weeks. Detoxification was done using benzodiazepines via a symptom-triggered regimen. Baclofen was offered post-detoxification with regular follow-ups. EEG signals were analysed for changes across standard frequency bands in various scalp regions and in terms of absolute powers.

Results

Sixty patients completed the 12-week follow-up. The sample consisted of all males, with a mean age of 40.85 ± 8.30 years. Alpha and gamma powers showed increasing trends, while beta, theta, and delta powers declined across most scalp regions. Absolute power trends were similar, with a statistically significant reduction noted in the delta wave (p-value=.024). No significant correlation was found between severity of dependence and wave powers, except for gamma in the fronto-parietal region (p-value=.01) and beta in the central region (p-value=.021).

Conclusion

qEEG changes with detoxification and abstinence may serve as an objective indicator for assessing treatment efficacy and abstinence status in alcohol dependence.

Keywords

qEEG electrophysiology benzodiazepines alcohol dependence withdrawal

Introduction

Alcohol, long used for social and recreational purposes, remains one of the leading causes of preventable morbidity and mortality worldwide. According to the World Health Organization (2018), approximately three million deaths annually are attributable to alcohol use, with nearly 14% occurring among individuals aged 20–40 years.¹ The neurobiological progression from casual consumption to alcohol dependence involves dysregulation of key neurotransmitter systems, including dopaminergic and γ-aminobutyric acid (GABA) pathways, glutamatergic hyperactivity, and disturbances within stress-related circuits.²

The International Classification of Diseases, 11th Revision (ICD-11) defines Alcohol Dependence Syndrome (ADS; 6C40.2) as a pattern of frequent and compulsive alcohol use characterised by impaired control, prioritisation of drinking, and persistence despite adverse consequences.³ Diagnosis typically requires the presence of symptoms over a 12-month period or nearly daily use for at least one month. Individuals with ADS commonly experience withdrawal symptoms within 8–24 h after cessation, with approximately 2%–8% developing hallucinations, 5%–10% progressing to delirium tremens, and about 10% experiencing seizures.^4–7

Treatment of ADS typically commences with detoxification, during which benzodiazepines (BDZs) remain the gold standard for managing withdrawal symptoms such as anxiety, tremors, and seizures.⁸ Two principal regimens are employed: the loading dose method and the symptom-triggered approach.⁹ The loading method involves administering long-acting BDZs to ensure gradual symptom resolution, though it carries risks such as oversedation, respiratory depression, and prolonged hospitalisation.¹⁰ The symptom-triggered approach, guided by the Clinical Institute Withdrawal Assessment for Alcohol–Revised (CIWA-Ar) scale, offers greater safety and shorter treatment duration but may lead to breakthrough symptoms or BDZ dependence.^9–11

Electroencephalography (EEG), traditionally used for assessing seizure disorders, has gained increasing prominence in psychiatry as an objective tool for studying brain function.¹² Quantitative EEG (qEEG) extends conventional EEG by converting raw electrical signals into numerical data that allows advanced spectral and coherence analyses.^13,14 Chronic alcohol consumption induces neuroadaptive changes in cortical neurons, reflected in both clinical manifestations and EEG alterations.¹⁵ Power spectral analysis (PSA), which quantifies electrical activity within specific frequency bands, provides insights into cortical dynamics and can differentiate alcohol-dependent individuals from healthy controls.¹⁶

Emerging evidence suggests that qEEG may also help predict the risk of relapse. Studies indicate that individuals who relapse often display persistent central nervous system (CNS) hyperarousal, marked by elevated beta activity, compared to those who remain abstinent.¹⁷ However, recent reviews have cautioned that the predictive validity of EEG measures remains inconsistent, emphasizing the need for longitudinal and standardized investigations.¹⁶

Mumtaz et al. demonstrated that EEG-derived features such as interhemispheric coherence and spectral power across delta, theta, alpha, beta, and gamma bands could accurately differentiate individuals with alcohol use disorder (AUD) from controls, achieving up to 89.3% classification accuracy.¹⁸ Bauer reported elevated high-frequency beta activity as a predictor of relapse, implicating frontal cortical dysregulation.¹⁹ Singh et al. observed increased beta and reduced delta/theta power in recently detoxified patients, suggesting sustained hyperarousal and vulnerability to relapse.²⁰ Jurado-Barba et al. further highlighted the clinical utility of EEG, identifying reduced alpha power as a marker of impaired cortical inhibition and variable theta oscillations associated with relapse risk or cortical dysfunction.²¹

Collectively, prior studies demonstrate consistent EEG abnormalities in alcohol-dependent populations—namely, increased beta and gamma activity reflecting hyperexcitability, and decreased alpha and delta rhythms indicating impaired inhibitory control.^22–24 These electrophysiological alterations are considered potential biomarkers of dependence severity and recovery trajectories.

Despite extensive international research, the clinical application of qEEG in alcohol dependence remains limited in India. Given the distinct sociocultural and biological determinants of alcohol use within this population, region-specific evidence is essential. Notably, there is a scarcity of longitudinal Indian studies examining qEEG changes before and after detoxification and during early abstinence. The present study addresses this gap by assessing pre- and post-treatment qEEG alterations in persons with Alcohol Dependence Syndrome, aiming to evaluate its potential as an objective biomarker for monitoring treatment response and neurophysiological recovery.

Methods

Description

Participants

A total of 80 consecutive patients diagnosed with Alcohol Dependence Syndrome according to ICD-11³ visiting the outpatient department of the department of Psychiatry of a tertiary care teaching hospital of North India were screened. Male patients in the age group of 25years and above, CIWA-Ar score <15, willing to provide informed consent and willing to get admitted to the deaddiction ward were enrolled for the study. Patients having epilepsy, dependence on other substances except nicotine and caffeine, past history of complicated withdrawals, currently intoxicated, having comorbid active psychotic illness, actively suicidal patients and with any acute medical or surgical illness were excluded from the study.

Out of 80 patients screened, 11 patients did not give consent, and 9 patients dropped out (4 due to loss to follow-up, 5 resumed dependent alcohol use), leaving 60 for final analysis.

Procedure

A semi-structured proforma was used to record socio-demographic and clinical data. Baseline investigations included a complete hemogram (CBC), liver function tests (LFT), serum electrolytes, and renal function tests (SERFT), a lipid profile, and an Electrocardiogram. At baseline, CIWA-Ar and the Severity of Alcohol Dependence Questionnaire (SADQ) scales were applied, followed by qEEG to assess correlation with alcohol dependence severity. Detoxification was done using lorazepam or diazepam as per the clinician's choice in a symptom-triggered regimen : lorazepam (4 mg/day initiation, 1 mg/day titration) and diazepam (10 mg/day initiation, 5 mg/day titration) orally, adjusted as per CIWA-Ar scores and side effect profiles. CIWA-Ar was recorded daily before the morning dose to monitor withdrawal symptoms by the principal investigator. Post-detoxification (CIWA-Ar <8), qEEG was repeated to evaluate changes in alpha, beta, delta, theta, and gamma wave activity as compared to baseline by the investigator. The post-detoxification EEG was recorded within 24–48 h after achieving CIWA-Ar < 8 and at least 24 h after the last benzodiazepine dose, ensuring that recordings reflected post-withdrawal neural activity rather than acute medication effects. Regular benzodiazepine use was stopped after detoxification, though occasional SOS use during follow-up was allowed.

All patients were offered Baclofen (10 mg/day) post-detoxification. In cases of baclofen use, dosing was up-titrated by 10 mg every 3 days based on cravings, unless adverse effects occurred, in which case adjustments were made more slowly (10 mg every 4-10 days or in 5 mg increments).²⁵

Further, patients were followed up regularly till 12 weeks of study (including the treatment with baclofen) to ensure compliance and abstinence. During this time period, if patients resorted to alcohol use again apart from dependent use, alcohol use pattern was recorded at follow ups. Moderate drinking or non-dependent use of alcohol was referred to as less than 14 drinks per week.²⁶ Intermittent users were patients who consumed <20% of their usual intake per day, not on three consecutive days, and not having any socio-occupational dysfunction due to alcohol use in the past 1 week.²⁷ For the purpose of this study, if patient had not consumed alcohol in 2 weeks prior to the date of follow up, it was considered as abstinence and if patient resumed drinking pattern as before he sought treatment, it was referred to as syndromal relapse. This operational definition was chosen to ensure EEG recordings were obtained under a physiologically stable abstinent condition while maintaining feasibility in real-world clinical settings, where sustained continuous abstinence for several months is uncommon.

Regular psychoeducation was provided and final qEEG assessment at 12 weeks was done along with application of scales and routine investigations. Comparison of overall change of qEEG parameters was done.

EEG Acquisition

The investigator received one month of formal EEG training under the Department of Medicine, under supervision of the EEG in-charge after which patient recruitment began. Each patient was informed about the procedure and instructed to remain awake and alert during the EEG. Recordings were conducted by the investigator in a soundproof room in the psychiatry ward for six minutes using the International 10–20 system with 18 electrodes (Fp1, Fp2, F7, F3, Fz, F4, F8, T7, C3, Cz, C4, T8, P7, P3, Pz, P4, P8, O2). Although 19 electrodes were applied, data from O1 was excluded due to persistent artefacts observed across recordings during the eyes-open condition despite repeated impedance checks; interpolation was not performed to avoid introducing reconstructed signal values, and the contralateral occipital electrode (O2) was retained to represent occipital activity. The 10–20 system ensures even electrode placement across the scalp, with electrode labels representing standard scalp regions (F = Frontal, P = Parietal, T = Temporal, O = Occipital) and numbers indicating laterality (odd = left, even = right, z = midline). In this study, terms such as “frontal,” “parietal,” and “occipital” regions refer to the topographical grouping of scalp electrodes according to the 10–20 system and are not intended to imply direct correspondence with underlying cortical sources. EEG data from all scalp areas was analysed. Impedance was kept below 5 kΩ. EEG signals were sampled at 1024 Hz, with 24-bit resolution, a 0.1–70 Hz bandpass filter, and a 50 Hz notch filter. EEG signals were recorded using linked earlobe electrodes (A1–A2) as the reference, and during pre-processing the data were re-referenced to a common average reference to minimize spatial bias and improve comparability across scalp locations. Recorded EEG data were digitally stored and pre-processed for artefact correction and noise reduction prior to quantitative analysis.

EEG Analysis

The EEG data were recorded for alcohol dependent patients under the 6-min eyes-open condition. The data were subsequently exported to the MATLAB (MathWorks, Natick, MA, USA) platform for further analysis (Figure 1). Continuous six-minute EEG data were segmented into 2-s non-overlapping epochs following artefact correction. Power spectra were computed for each clean epoch using FFT(Fast Fourier Transform)and averaged across all retained epochs to derive mean spectral power for each frequency band and electrode. On average, 160–170 artefact-free epochs were obtained per participant, representing approximately 85%–90% of total data, ensuring reliable and stable power estimates. Epochs with extreme values, ie exceeding ± 75µV in any of the 18 channels at any time within the epoch, were rejected. Epochs with power spectra demonstrating muscle and eye activity were rejected after visual inspection and labelling of components by ICA (Independent Component Analysis). Artefact free EEG signal was divided into frequency bands, to extract delta (0.5-4 Hz), theta (4-8 Hz), alpha (8-13 Hz), beta (13-30 Hz) and gamma (30-80 Hz) bands using Fast Fourier Transform (FFT) in all the 18 channels for baseline, after detoxification and at 12 weeks.

Ethical Considerations

The purpose and design of the study were explained to the patient in a manner they could understand. The patient had the right to withdraw from the study at any time without providing any reason. The confidentiality of the information gathered was maintained. The study was conducted in accordance with the defined guidelines of the Central Ethics Committee and the principles enunciated in the Declaration of Helsinki. The study was approved by the Institutional Ethics Committee and registered with the Clinical Trial Registry-India.

Statistical Analysis

Data was coded and recorded in MS Excel spreadsheet program. SPSS v23 (IBM Corp.)²⁸ was used for data analysis. Descriptive statistics were elaborated in the form of means/standard deviations for continuous variables, and frequencies/percentages for categorical variables. To enhance comparability across measurements, the min-max scaling method was used for standardisation of EEG data. Repeated-measures ANOVA was used to explore the change in qEEG over time. To control for potential inflation of Type I error due to multiple comparisons across frequency bands, the Bonferroni correction was applied. The assumption of normality was verified using the Shapiro–Wilk test, and sphericity was assessed using Mauchly's test. When sphericity was violated, Greenhouse–Geisser corrections were used to adjust the degrees of freedom. A p-value of less than .05 was considered statistically significant, and a p-value of less than .001 was considered highly significant.

Results

Socio-Demographic and Clinical Profile

The majority of participants were in the 36–45 age group (45%), predominantly married (86.7%), and from urban localities (66.7%). Most belonged to nuclear families (63.3%) and were educated up to matric (28.3%) or intermediate/diploma level (21.7%). Occupations were diverse, with the largest group being in the clerk/shop/farmer category (25%). Regarding alcohol use characteristics, more than half had a total duration of intake (TDI) between 11 and 20 years (51.7%) and a duration of dependent use between 0.5 and 6 years (91.6%). Most had average daily alcohol intake between 157–312 grams (53.3%), and recent intake mirrored this trend. Indian Made Foreign Liquor (IMFL) was the predominant type consumed (66.7%) as depicted in Table 1.

Table 1.

Sociodemographic and Clinical Profile of Participants of the Study.

		N = 60;n	%
Age in years [mean (SD)]		40.85 (8.30)
Marital Status	Unmarried	4	6.7%
	Separated	2	3.3%
	Divorced	2	3.3%
	Married	52	86.7%
Locality	Rural	20	33.3%
Locality	Urban	40	66.7%
Family Type	Nuclear	38	63.3%
	Extended	8	13.3%
	Joint	14	23.3%
Educational Qualification	Illiterate	4	6.7%
	Primary	6	10.0%
	Middle	11	18.3%
	Matric	17	28.3%
	Inter/Diploma	13	21.7%
	Graduate	9	15.0%
Occupation	Unemployed	11	18.3%
	Unskilled/Semi-Skilled	10	16.7%
	Skilled	12	20.0%
	Clerk/Shop/Farmer	15	25.0%
	Semi-Professional	7	11.7%
	Professional	5	8.3%
TDI (years)	0-10	17	28.3%
	11-20	31	51.7%
	21-30	10	16.7%
	31-40	2	3.3%
Non-Dependent Use (years)	0-10	40	66.7%
	11-20	18	30.0%
	21-30	1	1.7%
	31-40	1	1.7%
Dependent Use (years)	0.5-3	35	58.3%
	3.5-6	20	33.3%
	6.5-9	4	6.7%
	9.5-12	1	1.7%
Average Intake (grams)	0-156	22	36.7%
	157-312	32	53.3%
	>312	6	10.0%
Last Intake (grams)	0-156	23	38.3%
	157-312	34	56.7%
	>312	3	5.0%
Type of Alcohol	CML	6	10.0%
	IMFL	40	66.7%
	CML/IMFL	14	23.3%
Baclofen Prescription	Yes	41	68.3%
Baclofen Prescription	No	19	31.7%

TDI- Total duration of illness; SD – Standard deviation, CML – Country-made liquor, IMFL – Indian made foreign liquor

qEEG Changes

Scaled power values (Table 2) represent normalized EEG spectral power across scalp regions to depict relative changes over time, whereas absolute power values (Table 3) denote unscaled spectral energy and were used for statistical comparison.

Table 2.

Changes in various Wave Powers Across Scalp Regions (Electrode Groupings) at Baseline, post-Detoxification, and After 3 Months of Treatment.

		Alpha					Beta
		Mean (SD)	95% CIs	df	F-value	p-value	Mean (SD)	95% CIs	df	F-value	p-value
Fronto Parietal	Baseline	0.507 (0.224)	(0.449,0.565)	1.869	6.653	.538 (0.211)	0.538 (0.211)	(0.484,0.593)	1.866	0.458	.62
	After Detox	0.545 (0.229)	(0.486,0.604)			.539 (0.216)	0.539 (0.216)	(0.483,0.595)
	After 3 Months	0.407 (0.189)	(0.358,0.456)			.502 (0.258)	0.502 (0.258)	(0.435,0.569)
Frontal	Baseline	0.54 (0.137)	(0.505,0.575)	1.9	0.183	.543 (0.155)	0.543 (0.155)	(0.502,0.583)	1.969	0.762	.467
	After Detox	0.525 (0.174)	(0.48,0.57)			.513 (0.141)	0.513 (0.141)	(0.477,0.55)
	After 3 Months	0.537 (0.129)	(0.504,0.57)			.509 (0.152)	0.509 (0.152)	(0.47,0.548)
Temporal	Baseline	0.482 (0.194)	(0.431,0.532)	1.939	10.754	.463 (0.19)	0.463 (0.19)	(0.414,0.512)	1.917	2.399	.098
	After Detox	0.48 (0.221)	(0.423,0.537)			.449 (0.202)	0.449 (0.202)	(0.397,0.502)
	After 3 Months	0.616 (0.187)	(0.567,0.664)			.524 (0.204)	0.524 (0.204)	(0.471,0.576)
Parietal	Baseline	0.507 (0.133)	(0.472,0.541)	1.964	0.344	.489 (0.124)	0.489 (0.124)	(0.457,0.521)	1.921	0.055	.941
	After Detox	0.526 (0.133)	(0.492,0.561)			.497 (0.153)	0.497 (0.153)	(0.458,0.537)
	After 3 Months	0.509 (0.146)	(0.471,0.547)			.493 (0.14)	0.493 (0.14)	(0.457,0.529)
Occipital	Baseline	0.49 (0.319)	(0.408,0.573)	1.967	0.679	.559 (0.295)	0.559 (0.295)	(0.483,0.635)	1.972	0.846	.43
	After Detox	0.509 (0.279)	(0.437,0.581)			.561 (0.276)	0.561 (0.276)	(0.49,0.632)
	After 3 Months	0.553 (0.323)	(0.469,0.636)			.5 (0.335)	0.5 (0.335)	(0.414,0.587)
		Gamma					Theta
		Mean (SD)	95% CIs	df	F-value	p-value	Mean (SD)	95% CIs	df	F-value	p-value
Fronto Parietal	Baseline	0.523 (0.239)	(0.461,0.585)	1.535	15.356	.0001***	0.503 (0.228)	(0.445,0.562)	1.932	4.316	.017*
	After Detox	0.532 (0.237)	(0.471,0.594)				0.544 (0.184)	(0.497,0.592)
	After 3 Months	0.698 (0.162)	(0.657,0.74)				0.445 (0.203)	(0.393,0.498)
Frontal	Baseline	0.514 (0.153)	(0.474,0.553)	1.92	0.453	.629	0.526 (0.168)	(0.482,0.569)	1.837	2.435	.097
	After Detox	0.535 (0.131)	(0.501,0.569)				0.496 (0.144)	(0.459,0.533)
	After 3 Months	0.51 (0.133)	(0.476,0.545)				0.47 (0.12)	(0.439,0.501)
Temporal	Baseline	0.485 (0.239)	(0.423,0.547)	1.953	3.07	.051	0.509 (0.23)	(0.45,0.569)	1.925	0.036	.961
	After Detox	0.479 (0.214)	(0.423,0.534)				0.499 (0.216)	(0.443,0.554)
	After 3 Months	0.57 (0.21)	(0.516,0.624)				0.503 (0.219)	(0.446,0.559)
Parietal	Baseline	0.491 (0.137)	(0.455,0.526)	1.976	0.036	.964	0.492 (0.136)	(0.457,0.528)	1.98	0.301	.739
	After Detox	0.484 (0.141)	(0.448,0.521)				0.49 (0.128)	(0.457,0.523)
	After 3 Months	0.49 (0.141)	(0.453,0.526)				0.507 (0.126)	(0.474,0.539)
Occipital	Baseline	0.585 (0.293)	(0.51,0.661)	1.9	2.696	.075	0.581 (0.299)	(0.503,0.658)	1.926	4.016	.022*
	After Detox	0.578 (0.281)	(0.505,0.65)				0.457 (0.283)	(0.383,0.53)
	After 3 Months	0.467 (0.342)	(0.378,0.555)				0.431 (0.318)	(0.349,0.513)

Absolute Power		Mean	SD	95% CIs	df	F-value	p-value
Alpha	Baseline	0.507	0.069	(0.49,0.525)	1.964	0.696	.498
	After Detox	0.517	0.068	(0.499,0.534)
	After 12 Weeks	0.522	0.071	(0.503,0.54)
Beta	Baseline	0.516	0.065	(0.499,0.532)	1.902	0.794	.449
	After Detox	0.503	0.079	(0.483,0.523)
	After 12 Weeks	0.499	0.079	(0.479,0.52)
Gamma	Baseline	0.51	0.075	(0.491,0.53)	1.922	0.551	.571
	After Detox	0.519	0.078	(0.499,0.54)
	After 12 Weeks	0.525	0.073	(0.506,0.544)
Theta	Baseline	0.512	0.072	(0.494,0.531)	1.987	2.850	.062
	After Detox	0.494	0.074	(0.474,0.513)
	After 12 Weeks	0.485	0.072	(0.466,0.503)
Delta	Baseline	0.52	0.081	(0.499,0.541)	1.987	3.882	.024*
	After Detox	0.494	0.071	(0.475,0.512)
	After 12 Weeks	0.481	0.081	(0.46,0.502)

(*) – p-value <.05; S.D – Standard deviation

Table 3.

Changes in Absolute Powers of EEG Frequency Bands Across various Time Points.

Absolute Power		Mean	SD	95% CIs	df	F-value	p-value
Alpha	Baseline	0.507	0.069	(0.49,0.525)	1.964	0.696	.498
	After Detox	0.517	0.068	(0.499,0.534)
	After 12 Weeks	0.522	0.071	(0.503,0.54)
Beta	Baseline	0.516	0.065	(0.499,0.532)	1.902	0.794	.449
	After Detox	0.503	0.079	(0.483,0.523)
	After 12 Weeks	0.499	0.079	(0.479,0.52)
Gamma	Baseline	0.51	0.075	(0.491,0.53)	1.922	0.551	.571
	After Detox	0.519	0.078	(0.499,0.54)
	After 12 Weeks	0.525	0.073	(0.506,0.544)
Theta	Baseline	0.512	0.072	(0.494,0.531)	1.987	2.850	.062
	After Detox	0.494	0.074	(0.474,0.513)
	After 12 Weeks	0.485	0.072	(0.466,0.503)
Delta	Baseline	0.52	0.081	(0.499,0.541)	1.987	3.882	.024*
	After Detox	0.494	0.071	(0.475,0.512)
	After 12 Weeks	0.481	0.081	(0.46,0.502)

(***) – p-value <0.001; (**) – p-value < 0.01; (*) – p-value<0.05; S.D – Standard deviation.

Repeated-measures ANOVA (Table 2) revealed significant changes in alpha band power in the fronto-parietal and temporal scalp regions over the course of treatment. In the fronto-parietal scalp region (p = .002), alpha power increased slightly post-detoxification but showed a marked reduction at the 3-month follow-up compared to baseline. Similarly, in the temporal scalp region (p < .001), alpha power remained nearly unchanged after detoxification but increased significantly at the 3-month mark.

For most scalp regions, beta wave power (Table 2) showed a declining trend over the course of abstinence, except in the temporal and parietal areas. However, Repeated-Measures ANOVA indicated that these changes were not statistically significant in any of the regions.

Repeated-Measures ANOVA also revealed significant changes in theta band power (Table 2) in the fronto-parietal (p = .017) and occipital (p = .022) scalp regions. In the fronto-parietal area, theta power increased following detoxification but decreased at the 3-month follow-up compared to baseline. In the occipital scalp region, there was a consistent and significant decline in theta power from baseline to 3 months.

Repeated-Measures ANOVA further showed a statistically significant increase in gamma band power (Table 2) in the fronto-parietal scalp region (p < .001), with a modest rise post-detoxification followed by a more pronounced elevation at the 3-month follow-up. Gamma activity in the temporal scalp region (p = .051) also showed a steady upward trend during abstinence.

Repeated-Measures ANOVA did not show any statistically significant changes in delta band power (Table 2) across scalp regions during the treatment period, although minor nonsignificant decreases were noted over time.

Repeated-measures ANOVA (Table 3) demonstrated that, among all frequency bands, only the absolute power of delta waves showed a statistically significant reduction over time (p = .024*), with values decreasing from baseline to 12 weeks.

Clinical Assessment Scales

There was a highly significant reduction in CIWA-Ar scores over time (p < .001), with mean scores dropping from 10.98 at baseline to 1.97 after detoxification, and further declining to near zero (0.07) at 12 weeks. This indicates effective management of withdrawal symptoms throughout the treatment period. The correlation between absolute change in qEEG frequency bands from baseline to endline and change in CIWA-Ar scores was examined. Among the frequency bands, beta activity showed a weak positive correlation with CIWA-Ar change (r = 0.224), approaching statistical significance (p = .085), suggesting a possible trend toward higher beta activity being associated with greater withdrawal symptom reduction. However, this result did not reach statistical significance. All other frequency bands — alpha (r = 0.029, p = .826), gamma (r = -0.038, p = .775), theta (r = 0.089, p = .498), and delta (r = 0.073, p = .580) — showed very weak and statistically non-significant correlations with CIWA-Ar change. Overall, no significant associations were observed between changes in qEEG spectral power and withdrawal severity as measured by CIWA-Ar.

The mean SADQ score was 22.6 (SD = 4.8), indicating moderate alcohol dependence severity. At baseline, the SADQ showed a significant positive correlation with gamma power in the fronto-parietal region (r = 0.332, p = .01), indicating that higher severity of dependence is associated with increased gamma activity in this area. Additionally, beta power in the central region correlated positively with SADQ scores (r = 0.297, p = .021). The occipital theta band showed a trend toward significance (r = 0.253, p = .051)

Discussion

The present study aimed to examine longitudinal changes in quantitative EEG (qEEG) parameters in patients with alcohol dependence syndrome (ADS) over 12 weeks of treatment and to explore their association with baseline severity of dependence (SADQ scores).

The study's participants (Table 1) were primarily middle-aged married males, consistent with previous studies on socio-demographic profiles of alcohol dependent patients by Vignesh et al., 2014²⁹ and Gupta et al., 2020.³⁰ Educational backgrounds varied across participants, and the findings were consistent with those of Pradeep et al. (2010).³¹ Conversely, Vignesh et al.²⁹ noted a significant association between lower education and dependence. The occupation, residence and type of families of participants also echoed findings from a previous study by Vignesh et al.²⁹ Clinically, most participants had consumed alcohol for 11–20 years, paralleling durations reported by Manning et al.,³² though shorter than the 44 years reported by Munro et al.,³³ which could be due to different age demographics. The non-dependent use phase and dependence phase were shorter than the durations reported in studies by Pitel et al.,³⁴ Zinn et al.,³⁵ and Brion et al.³⁶ These variations may be explained by genetic differences (eg, ADH1B and ALDH2 variants) across ethnic populations acting as protective factors.³⁷ IMFL was the most consumed alcohol type, likely due to urban accessibility. Baclofen was prescribed in most cases, supporting its documented effectiveness in reducing cravings by Abhijit et al.²⁷

Absolute alpha power (Table 3) and in temporal, parietal, and occipital scalp regions (Table 2) was seen to have an increasing trend, aligning with findings by Mumtaz et al.³⁸ However, in frontal and fronto-parietal scalp regions, decreasing trends were noted, which suggests region-specific alterations in cortical activity.

Absolute beta power (Table 3) and across fronto-parietal, frontal, and occipital regions of the scalp (Table 2) decreased over time, which is consistent with the Rangaswamy et al.²¹ Study, which reported elevated beta activity in alcoholics that declined with abstinence. Declining trends support the hypothesis that beta power normalisation indicates neural stabilisation and recovery. Some regions (eg fronto-parietal, parietal and occipital regions) showed an initial post-detoxification increase before declining eventually, which was consistent with Mumtaz et al.¹²

Absolute theta power (Table 3) and in fronto-parietal, frontal, temporal and occipital regions of the scalp (Table 2) were reduced, aligning with the results of Rangaswamy et al.²³ Significant reductions were seen over fronto-parietal and occipital scalp regions, in contrast to the predominantly parietal theta abnormalities reported in earlier studies by Rangaswamy et al. (2003)²³ and Singh et al. (2018).²⁰ This may reflect differences in timing of EEG assessments. The reduction in theta activity suggests improved cortical efficiency and cognitive regulation during abstinence.^39,40

Absolute gamma power (Table 3) and in fronto-parietal and temporal regions (Table 2) increased over time, consistent with Porjesz et al.,²⁴ which observed lower gamma activity in alcohol-dependent individuals with recovery during abstinence. Gamma waves relate to working memory and attention, and an increase in value may indicate neuroplastic changes in pre-frontal networks during recovery.^41,42

Absolute delta power and in all regions of scalp (Tables 2 and 3) declined, matching the findings of Dijk et al.⁴³ but conflicting with Colrain et al.⁴⁴ who found reduced delta power in dependence. Given delta's association with deep sleep and brain recovery, these changes may reflect differential recovery patterns across cortical regions during abstinence.

Observed inconsistencies with prior literature may be attributed to small sample size, inter-individual variability, ethnic factors, and the limited 12-week duration. Nevertheless, the overall qEEG trends support existing neurobiological models of alcohol dependence and recovery.

In the present study, analysis of EEG parameters in relation to the SADQ scores at baseline reveals no significant correlations except gamma power in fronto-parietal region and beta power in central region. The limited association could possibly hint at underlying neural patterns that may not be reflected in clinical scales. To the best of our knowledge, no study has previously compared these parameters and thus the findings could not be corroborated with existing literature.

These region-specific changes further underscore the need to explore how distinct cortical networks respond differentially to treatment and recovery.

Strengths and Limitations

This study had several strengths, including a robust prospective, randomized, interventional design and the use of validated clinical scales (CIWA-Ar, SADQ) for comprehensive assessment. Inclusion of baclofen and psychoeducation reflected real-world treatment strategies, supporting abstinence and adherence. A three-month (12 week) follow-up allowed for evaluation of both immediate and sustained effects, while EEG data were analysed using precise MATLAB-based techniques with artefact rejection to enhance accuracy. However, the study had limitations, including a small sample size (n = 60) and a male-only sample, limiting generalizability. Being a single-center study also reduced external validity. Analyses were based on complete cases, and although attrition was low, minor bias due to non-random missing data cannot be entirely ruled out. The lack of a non-medication control group made it difficult to differentiate treatment effects from natural recovery. Use of different benzodiazepines and baclofen in some cases reflects routine clinical practice but may have introduced minor variability. Baclofen, though offered as part of standard care and used by most participants, may have had minor unmeasured effects on qEEG parameters that could not be fully controlled for. Occasional SOS use of benzodiazepines during follow-up could not be entirely excluded. While the three-month follow-up offered insight into short-term outcomes, it may not reflect long-term recovery. Self-reported data on craving, and abstinence were subject to recall and social desirability bias. Additionally, variability in the duration of abstinence – defined as at least two weeks without alcohol before follow-up – introduced inconsistency in interpreting qEEG findings. This pragmatic definition was adopted to ensure a stable physiological state for EEG recording while maintaining feasibility in real-world clinical settings, but it may not fully reflect continuous abstinence throughout the follow-up period.

Conclusion

This study identified absolute and region-specific qEEG changes during early abstinence in individuals with alcohol dependence, with significant alterations noted in alpha, theta, and gamma bands. Gamma activity in the fronto-parietal region and beta power in the central region showed correlations with severity of dependence. These findings suggest that qEEG may serve as a useful, non-invasive adjunct for objectively monitoring neurophysiological changes during treatment and early abstinence. However, given the modest sample size and absence of a control group, the results should be interpreted with caution. Future studies with larger cohorts, longer follow-up, and comparative designs are warranted to confirm these trends and explore their clinical relevance.

Footnotes

Acknowledgment

None

ORCID iD

Shivangi Mehta

Ethical Approval

The study was approved by the Institutional Ethics Committee.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Disclosures

None

References

Hyun

Han

Lee

Yoon

Jung

. Pathophysiological aspects of alcohol metabolism in the liver. Int J Mol Sci. 2021;22:5717. doi:10.3390/ijms22115717

Roberto

Gilpin

Siggins

. The central amygdala and alcohol: Role of γ-aminobutyric acid, glutamate, and neuropeptides. Cold Spring Harb Perspect Med. 2012;2:a012195. doi:10.1101/cshperspect.a012195

World Health Organization. International statistical classification of diseases and related health problems. 11th revision. Geneva: World Health Organization; 2019. Last accessed October 24, 2025. https://icd.who.int/.

Schuckit

Klein

. Correlations between drinking intensity and reactions to ethanol and diazepam in healthy young men. Neuropsychopharmacology. 1991;4:157–163.

Wood

Albarqouni

Tkachuk

, et al. Will this hospitalized patient develop severe alcohol withdrawal syndrome? The rational clinical examination systematic review. JAMA. 2018;320:825–833. doi:10.1001/jama.2018.10574

Stephane

Arnaout

Yoon

. Alcohol withdrawal hallucinations in the general population, an epidemiological study. Psychiatry Res. 2018;262:129–134. doi:10.1016/j.psychres.2018.02.021

Hillbom

Pieninkeroinen

Leone

. Seizures in alcohol-dependent patients: Epidemiology, pathophysiology and management. CNS Drugs. 2003;17:1013–1030. doi:10.2165/00023210-200317140-00002

Ozdemir

Bremner

Naranjo

. Treatment of alcohol withdrawal syndrome. Ann Med. 1994;26:101–105. doi:10.3109/07853899409147336

Sellers

Naranjo

Harrison

Devenyi

Roach

Sykora

. Diazepam loading: Simplified treatment of alcohol withdrawal. Clin Pharmacol Ther. 1983;34:822–826. doi:10.1038/clpt.1983.256

10.

Maldonado

DiMartini

Owen

. Psychopharmacological treatment of substance use disorders in the medically ill. In: Levenson

Ferrando

, eds. Clinical Manual of Psychopharmacology in the Medically Ill. American Psychiatric Publishing; 2010:537–555.

11.

McKay

Koranda

Axen

. Using a symptom-triggered approach to manage patients in acute alcohol withdrawal. Medsurg Nurs. 2004;13:15–20.

12.

Mumtaz

Vuong

Malik

Rashid

RBA

. A review on EEG-based methods for screening and diagnosing alcohol use disorder. Cogn Neurodyn. 2018;12:141–156. doi:10.1007/s11571-017-9465-x

13.

Klonowski

Jernajczyk

Niedzielska

Rydz

Stepień

. Quantitative measure of complexity of EEG signal dynamics. Acta Neurobiol Exp (Wars). 1999;59:315–321. doi:10.55782/ane-1999-1316

14.

Straaten

ECW

Stam

. Structure out of chaos: Functional brain network analysis with EEG, MEG, and functional MRI. Eur Neuropsychopharmacol. 2013;23:7–18. doi:10.1016/j.euroneuro.2012.10.010

15.

Coutin-Churchman

Moreno

Añez

Vergara

. Clinical correlates of quantitative EEG alterations in alcoholic patients. Clin Neurophysiol. 2006;117:740–751. doi:10.1016/j.clinph.2005.12.021

16.

Zhang

Yao

Wang

. Targeting electroencephalography for alcohol dependence: A narrative review. CNS Neurosci Ther. 2023;29:1205–1212. doi:10.1111/cns.14138

17.

Saletu-Zyhlarz

Arnold

Anderer

, et al. Differences in brain function between relapsing and abstaining alcohol-dependent patients, evaluated by EEG mapping. Alcohol Alcohol. 2004;39:233–240. doi:10.1093/alcalc/agh041

18.

Mumtaz

Vuong

Xia

Malik

Rashid

RBA

. An EEG-based machine learning method to screen alcohol use disorder. Cogn Neurodyn. 2017;11:161–171. doi:10.1007/s11571-016-9416-y

19.

Bauer

. Predicting relapse to alcohol and drug abuse via quantitative electroencephalography. Neuropsychopharmacology. 2001;25:332–340. doi:10.1016/S0893-133X(01)00236-6

20.

Singh

Nizamie

Tikka

. Sleep architecture and EEG power spectra in recently detoxified alcohol dependent patients. Asian J Psychiatr. 2018;32:126–136. doi:10.1016/j.ajp.2017.12.005

21.

Jurado-Barba

Sion

Martínez-Maldonado

, et al.

Neuropsychophysiological measures of alcohol dependence: Can we use EEG in the clinical assessment?

Front Psychiatry. 2020;11:676. doi:10.3389/fpsyt.2020.00676

22.

Rangaswamy

Porjesz

Chorlian

, et al. Beta power in the EEG of alcoholics. Biol Psychiatry. 2002;52:831–842. doi:10.1016/S0006-3223(02)01362-8

23.

Rangaswamy

Porjesz

Chorlian

, et al. Theta power in the EEG of alcoholics. Alcohol Clin Exp Res. 2003;27:607–615. doi:10.1111/j.1530-0277.2003.tb04397.x

24.

Porjesz

Begleiter

. Alcoholism and human electrophysiology. Alcohol Res Health. 2003;27:153–160.

25.

Imbert

Blaise

Bérard

de Beaurepaire

Stafford

Jaury

. A prescription guide for baclofen in alcohol use disorder- for use by physicians and patients. J Addict Med Ther Sci. 2017;3:32–41.

26.

NIAAA. Understanding alcohol drinking patterns. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2023. Last accessed March 25, 2025. https://www.niaaa.nih.gov/alcohols-effects-health/alcohol-drinking-patterns.

27.

Rozatkar

Kapoor

Sidana

Chavan

. Clinical experience of baclofen in alcohol dependence: A chart review. Ind Psychiatry J. 2016;25:11–16. doi:10.4103/0972-6748.196043

28.

IBM Corp. IBM SPSS Statistics for Windows, Version 23.0. IBM Corp; 2015.

29.

Vignesh

Singh

Mohan

Murthy

Joshi

. Association between socio-demographics and alcohol dependence among individuals living in an Indian setting. Glob J Health Sci. 2014;6:16–26. doi:10.5539/gjhs.v6n3p16

30.

Gupta

Sinha

Khanna

Kumar

Sharma

. Marital status and alcohol dependence: A clinical perspective. Indian J Psychiatry. 2020;62:297–304.

31.

Pradeep

Banu

Ashok

. Severity of alcoholism in Indian males: Correlation with age of onset and family history of alcoholism. Indian J Psychiatry. 2010;52:243–249. doi:10.4103/0019-5545.70977

32.

Manning

Wanigaratne

Best

, et al. Changes in neuropsychological functioning during alcohol detoxification. Eur Addict Res. 2008;14:226–233. doi:10.1159/000156479

33.

Munro

Saxton

Butters

. The neuropsychological consequences of abstinence among older alcoholics: A cross-sectional study. Alcohol Clin Exp Res. 2000;24:1510–1516.

34.

Pitel

Rivier

Beaunieux

Vabret

Desgranges

Eustache

. Changes in the episodic memory and executive functions of abstinent and relapsed alcoholics over a 6-month period. Alcohol Clin Exp Res. 2009;33:490–498. doi:10.1111/j.1530-0277.2008.00859.x

35.

Zinn

Stein

Swartzwelder

. Executive functioning early in abstinence from alcohol. Alcohol Clin Exp Res. 2004;28:1338–1346. doi:10.1097/01.ALC.0000139814.81811.62

36.

Brion

D’Hondt

Pitel

, et al. Executive functions in alcohol-dependence: A theoretically grounded and integrative exploration. Drug Alcohol Depend. 2017;177:39–47. doi:10.1016/j.drugalcdep.2017.03.018

37.

Edenberg

. The genetics of alcohol metabolism: Role of alcohol dehydrogenase and aldehyde dehydrogenase variants. Alcohol Res Health. 2007;30:5–13.

38.

Mumtaz

Vuong

Xia

Malik

Rashid

RBA

. Automatic diagnosis of alcohol use disorder using EEG features. Knowl Based Syst. 2016;105:48–59. doi:10.1016/j.knosys.2016.04.026

39.

Klimesch

. EEG Alpha and theta oscillations reflect cognitive and memory performance: A review and analysis. Brain Res Rev. 1999;29:169–195. doi:10.1016/S0165-0173(98)00056-3

40.

Knyazev

. Motivation, emotion, and their inhibitory control mirrored in brain oscillations. Neurosci Biobehav Rev. 2007;31(3):377–395. doi:10.1016/j.neubiorev.2006.10.004

41.

Başar-Eroglu

Başar

Demiralp

Schürmann

. P300-response: Possible psychophysiological correlates in delta and theta frequency channels. Int J Psychophysiol. 1996;24(1-2):1–21.

42.

Porjesz

Rangaswamy

Begleiter

. The genetics of brain oscillations. Int J Psychophysiol. 2005;58(2-3):179–196.

43.

Dijk

Brunner

Aeschbach

Tobler

Borbély

. The effects of ethanol on human sleep EEG power spectra differ from those of benzodiazepine receptor agonists. Neuropsychopharmacology. 1992;7:225–232.

44.

Colrain

Turlington

Baker

. Impact of alcoholism on sleep architecture and EEG power spectra in men and women. Sleep. 2009;32:1341–1352. doi:10.1093/sleep/32.10.1341