Morphological Changes of the Retina in Alcohol Use Disorder: A Systematic Review and Meta-analysis of Studies Using Optical Coherence Tomography

Abstract

Purpose of the Review:

Chronic alcohol use is associated with various structural and functional changes in the brain. Retinal morphology assessed by optical coherence tomography (OCT) non-invasively detects alcohol related damage to the brain and can be a disease marker.

Collection and Analysis of Data:

A systematic review of studies comparing retinal morphology using OCT, between alcohol use disorder (AUD) patients and healthy controls (HC) from PubMed, Scopus and Embase databases was performed (on 15/April/2025). Random effects meta-analyses were conducted for the thickness of retinal parameters, at both disc and macula: Retinal nerve fibre layer (RNFL), ganglion cell layer (GCL), inner plexiform layer (IPL), ganglion cell-inner plexiform layer (GCIPL), choroid thickness (CT), macular thickness (MT) and macular volume (MV). The Newcastle-Ottawa Scale was used for risk of bias (RoB) assessment. Publication bias, sensitivity analysis and certainty of evidence (CoE) was assessed using Doi plots, the leave-one-out method and the GRADE approach, respectively.

Results:

Of the 2,416 records screened, eight studies (n = AUD = 6,276 eyes; HC = 2,695 eyes) were included in meta-analyses. They revealed significant thinning of the total (pooled SMD = −0.41; 95% CI = −0.68, −0.14; I² = 75%; k = 6) and nasal (pooled SMD = −0.36; 95% CI = −0.58, −0.13; I² = 56%; k = 6) peripapillary RNFL in AUD patients. Significantly lower average MT (pooled SMD = −0.62; 95% CI = −0.95, −0.29; I² = 50%; k = 3) and macular GCIPL thickness (pooled SMD = −0.19; 95% CI = −0.33, −0.06; I² = 67%; k = 3) were shown. CoE was ‘moderate’ for total peripapillary RNFL thinning, but was ‘very low’ for other outcomes, owing to heterogeneity and publication bias. RoB assessment showed one study with unsatisfactory quality.

Conclusions:

Evidence for thinning of retinal layers, especially the peripapillary RNFL, as AUD disease-markers is promising, but preliminary. Our results align with the hypothesis that chronic alcohol consumption induces neurodegenerative changes in the retina and, therefore, the brain.

Keywords

Addiction alcohol dependence macular thickness retinal morphology retinal nerve fibre layer thickness substance use disorders

Alcohol use disorder (AUD) is widely prevalent in the world and is associated with significant morbidity and mortality. Prolonged alcohol use tends to provoke or exacerbate both psychiatric disorders and neurological disorders such as stroke, multiple sclerosis and dementia.¹ Structural changes such as global cerebral, cerebellar atrophy and central pontine myelinolysis are known to be caused by chronic alcohol use. The primary pathophysiology associated with these neurological sequelae is neurodegeneration, neuro-immune alterations and irreversible brain injury.² Perhaps, even in the absence of such overt sequelae, chronic use of alcohol is hypothesised to cause neuronal damage by way of oxidative stress, neuroinflammation and cellular injury. Several animal studies have shown that alcohol induces neuroinflammation, that is, induction of pro-inflammatory cytokines and chemokines in the brain.³

Optical coherence tomography (OCT) acts as a rapid and non-invasive technique for assessing retinal morphology and architecture, which are proxy markers of pathophysiological changes happening in the brain. The technique uses infrared electromagnetic waves to provide cross-sectional images of tissues. It has been widely used for diagnosing and monitoring ocular pathologies.⁴ Use of OCT in understanding the pathogenesis of various psychiatric disorders, particularly mood and psychotic disorders, has been demonstrated in recent systematic reviews and meta-analyses.^5–6

Research related to OCT findings in substance use disorders is on the rise. The research so far does show that alcohol induced retinal changes can act as a non-invasive window to the changes that happen in the brain. In one of the earlier studies, Ahuja et al.⁷ demonstrated a significant reduction in retinal nerve fibre thickness (RNFL) found in all quadrants except nasal in patients with AUDs compared to healthy controls (HC). Subsequently, there have been some contrasting findings too—for example, Özsoy and Alim⁸ showed that patients with AUD did not differ from controls in any of the retinal parameters on OCT. To date, there has been no synthesis of findings on this aspect. Hence, we aimed to perform a systematic review and meta-analysis on studies evaluating retinal morphology, using OCT both at disc and macula in patients with AUDs and assessing its disease marker status.

Methods

We carried out a systematic review and meta-analysis focusing on studies that assessed retinal morphology in patients with AUD. A search conducted in Cochrane and The International Prospective Register of Systematic Reviews (PROSPERO) databases before conducting the review found no similar systematic review and meta-analysis. The study protocol was registered in PROSPERO. The review cum meta-analysis was carried out as per the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) extension guidelines (https://www.prisma-statement.org/). The study was conducted between August 2024 and June 2025.

Selection Criteria

Documents meeting our inclusion criteria for the review were all published article types (except case reports/series) with original data, provided sufficient data required for synthesis is available. Studies included patients with AUD or alcohol dependence, diagnosed according to standard classificatory systems. Studies where participants were defined to have ‘regular’ or ‘daily’ consumption of alcohol were also included. Assessment of retinal morphology should have been carried out using OCT exclusively. The population could be of any age and sex.

Search Strategy

A thorough search was conducted in PubMed, Embase and Scopus databases for all articles published till 15 April 2025 using the search string: (‘Alcohol*’[Title/Abstract] OR ‘alcohol*’[MeSH Terms]) AND (‘optical coherence tomogra*’[Title/Abstract] OR ‘ocular coherence tomogra*’[Title/Abstract] OR ‘retina* nerve’[Title/Abstract] OR ‘retina* morphology’[Title/Abstract] OR ‘macula* thick*’[Title/Abstract] OR ‘choroid* thick*’[Title/Abstract] OR ‘fovea*’[Title/Abstract] OR ‘fovea*’[MeSH Terms] OR ‘maxwell spot’[Title/Abstract] OR ‘ganglion* cell layer’[Title/Abstract] OR ‘Plexiform Layer’[Title/Abstract] OR ‘vascular coat’[Title/Abstract] OR ‘Nuclear Layer’[Title/Abstract] OR ‘Photoreceptor Layer’[Title/Abstract] OR ‘Photoreceptor Layer’[Title/Abstract] OR ‘Soemmering ring’[Title/Abstract]). Details of the PubMed search history are provided in Supplementary Table S1. The ‘query translator’ feature was used to translate the search string from PubMed to Embase. A ‘title, abstract and keyword’ search on Scopus was carried out using the same search terms.

Table 1.

Summary of Findings of the Meta-analyses on OCT in Patients with AUD and HC.

Retinal Layer	K/N			Main Effects			Heterogeneity
Retinal Layer	Studies	AUD Eyes/Pts	HC Eyes/ Pts	SMD (AUD vs HC)	95% CI	ρ	τ²	ι²	χ²
Disc
Superior pRNFL	6	442/257	496/284	−0.21	−0.53, 0.10	.180	0.17	78%	36.86*
Inferior pRNFL	6	442/257	496/284	−0.24	−0.51, 0.02	.07	0.11	70%	26.50*
Temporal pRNFL	6^#/5	400/236	454/263	−0.25	−0.76, 0.27	.35	0.43	91%	66.56*
Nasal pRNFL	6	442/257	496/284	−0.36	−0.58, −0.13	.002	0.06	56%	18.36
Total pRNFL	6	829/744	959/947	−0.41	−0.68, −0.14	.003	0.11	75%	31.49*
GCIPL, GCL, IPL	Not provided in any of the included studies
pOPL; pONL; pRPE; pINL	Not provided in any of the included studies
Macula
mRNFL	2	5,289/5,247	1,578/1,557	−0.19	−0.50, 0.12	.23	0.04	44%	3.58
mGCL + IPL	3	5,877/5,877	2,242/2,242	−0.19	−0.33, −0.06	.004	0.01	67%	6.11
MT Central average	3 3	137/90	191/117	−0.24 −0.62	−0.51, 0.04 −0.95, −0.29	.09 .0002	0.03 0.07	31% 50%	5.78 8.03
mGCL	1	42/21	42/21	Could not be analysed as only one study provided data
mGCC	1	42/21	42/21	Could not be analysed as only one study provided data
mIPL	1	42/21	42/21	Could not be analysed as only one study provided data
mCT	2	119/ 81	119/81	Could not be computed as both studies provided different parameters (Özsoy: Nasal, subfoveal and temporal CT; Orum: Total CT)
mOPL; mONL; mRPE; mINL	Not provided in any of the included studies

pRNFL: Peripapillary RNFL, mRNFL: Macular RNFL, mGCL +IPL: Macular ganglion cell layer + inner plexiform layer, MT: Macular thickness, mGCC: Macular ganglion cell complex, IPL: Macular IPL, CT: Macular choroidal thickness, OPL: Macular outer plexiform layer, ONL: Macular outer nuclear layer and RPE: Macular retinal pigmented epithelium, INL: Inner nuclear layer, Alc eyes: Number of eyes of patients, HC eyes: Number of eyes of controls, SMD: Standardised mean difference, CI: Confidence interval, # –Alvarez et al. not included for meta-analysis as values are erroneous, x²: tau², |²: chi-square, *p < .05.

Study Screening

Screening was done using (https://new.rayyan.ai/). Duplicates were assessed and removed (by SKT). The initial screening was carried out by going through the titles and abstracts by two independent investigators (MSS and SKT). After the first level screening, the second level screening of the full text of ‘eligible articles’ was also done by the same authors (Table S2). Conflicts were resolved after discussion with the third investigator (SSJ). Cross-references were also looked for to complete the search process (See Figure 1).

Outcome Measures and Data Extraction

OCT parameters at both disc and macula: RNFL thickness, ganglion cell layer (GCL) thickness, inner plexiform layer (IPL) thickness, ganglion cell-inner plexiform layer (GCIPL) thickness, choroid thickness (CT), macular thickness (MT) and macular volume (MV) between patients with AUD and HC across the included studies were extracted. These measures, when available for other sub-specifiers such as right or left eye and quadrants or grids, were extracted along with total or average measures. Mean, standard deviation (SD) and sample size for each of these outcome measures were extracted. For measures such as superior RNFL and inferior RFNL, where the data were provided for two sub-quadrants (such as temporal-superior and nasal-superior), the combined mean and SD were calculated using an open-source platform meta-analysis accelerator.⁹

Risk of Bias Assessment

Risk of bias (RoB) assessment was carried out independently by two investigators (BCM and STT). Newcastle-Ottawa Scale¹⁰ (for cross-sectional studies) was used for quality assessment.

Statistical Analysis

All the outcome data from the studies selected for meta-analysis were entered into Microsoft Excel. This data was imported into RevMan software version 5.4.1¹¹ for conducting meta-analysis. Data were grouped by outcome and pooled using a random effects model (inverse variance method) based on standardised mean differences (SMD). SMD values of 0.2–0.5, 0.5–0.8 and >0.8 were considered small, medium and large, respectively. The level of heterogeneity was estimated using tau² (using the restricted maximum-likelihood estimator method), the Q-test for heterogeneity (p < .1) and the I² statistic (<25%: Low; 25%–50%: Moderate; >50%: High). Meta-analyses were conducted only where k ≥ 3.

Publication Bias, Sensitivity Analysis, Subgroup Analysis and Meta-regression

As the number of included articles was less than 10, Doi plots with Luis Furuya-Kanamori (LFK) index were used to assess the publication bias. The leave-one-out method was conducted for sensitivity analysis. For Doi plots with LFK index and leave-one-out analysis, the R language with R Studio¹² as Integrated Development Environment (IDE) and ‘meta’, ‘metasens’, ‘metafor’, ‘dplyr’ and ‘ggplot2’ packages were used. Subgroup analysis for all outcomes was carried out for eye-sidedness and OCT type (Table S4). Meta-regression was carried out using the ‘moderator’ function in JAMOVI¹³ (Major plugin) for age, duration of AUD and average daily amount of alcohol.

Certainty of Evidence (CoE)

Certainty of evidence (CoE) was assessed using GRADE-pro.¹⁴

Results

A total of 2,416 articles were identified. Thirty-one articles were retrieved after duplicates’ removal and initial screening. Two investigators (MSS and SKT) went through the full text and after mutual agreement, it was decided to include nine articles for final analysis. Of the nine articles, eight were included in the meta-analysis (Details of the 22 excluded studies are provided in Table S2). One study,¹⁵ with only six patients of AUD, was excluded as a detailed description of outcome measures was not provided. Details of screening and eligibility are provided in the PRISMA flow chart (See Figure 1).

Figure 1.

PRISMA Flow Chart.

Study Characteristics

Characteristics of the included studies are described in detail in Table S3. All the included studies were published between 2016 and 2024.^7,8,15–21 Sample size of AUD patients evaluated in the included studies ranged from 6 to 5,247 subjects.^7,8,15–21 However, for the studies included in meta-analysis, the sample size ranged from 21 to 5,247; for subjects with AUD, 42–5,247 eyes; and for HC, 42–1,536 eyes (21–1,536 subjects).^7,8,16–21 Eight out of nine studies used spectral domain OCT for the evaluation of retinal morphology; Ahuja et al.⁷ used time-domain OCT.

Risk of Bias Assessment

Except for one study,¹⁵ which showed unsatisfactory quality, all other studies were found to be of good to satisfactory quality (Table S4). This study was not included in the meta-analysis.

Qualitative Synthesis

Table S3 describes the results obtained by the included studies. Out of the nine included studies, one,²¹ did not report findings on peripapillary retinal nerve fibre thickness (pRNFL). While five,^7,17–20 of these eight report significant thinning of pRNFL in AUD patients compared to HC, one,⁸ report increased thickness and two,^15,16 report no differences. Differential thinning of subregions was noted. Liu et al.¹⁸ report significant thinning of pRNFL restricted to the left eyes and to the temporal and nasal quadrants. Specific thinning of pRNFL in temporal regions was found by Sahin et al.¹⁹ too. Interestingly, the Indian study by Ahuja et al.⁷ showed thinning of pRNFL in all subregions except the nasal quadrant. Five^{8,16,18,20–21} of the nine report findings on thickness at macula, including total (i.e., MT), central, that is, foveal, macular retinal nerve fibre thickness (mRNFL) and macular ganglion cell-inner plexiform layer (mGCIPL). Two studies, Álvarez-Sesmero et al.¹⁶ and Liu et al.¹⁸ report significant thinning in the total mean macular layer (i.e., MT) of AUD patients compared to HC. Liu et al.¹⁸ report significant thinning of the central macular (or foveal) layer, specifically in the right eye and Ho et al.²⁰ report significant thinning in the left mGCIPL. Özsoy and Alim⁸ found no differences between AUD patients and HC in any of the macular parameters, including the choroid layers.

Meta-analysis (Including Subgroup Analysis)

Total Peripapillary Retinal Nerve Fibre Thickness (pRNFL)

Total pRNFL was found to be significantly thinner in the AUD group than in the control group. (SMD = −0.41; 95% CI = −0.68, −0.14; k = 9; Z = 3.01, p < .05; I² = 75%). Overall subgroup differences based on sidedness were not significant (p = .20). Left-sided RNFL was found to be significantly thinner in the AUD group than the control group (p = .03) (Figure 2A).

Figure 2.

RNFL: Retinal nerve fibre layer, Alcohol: Alcohol use disorder group, Control: Healthy control group, SD: Standard deviation, IV: Inverse variance, CI: Confidence interval, Random: Random effects model, df: Degrees of freedom.

Superior Peripapillary Retinal Nerve Fibre Thickness (pRNFL)

No difference in the superior RNFL thickness was found between the AUD group and the control group. (SMD = −0.21; 95% CI = −0.53, 0.10; k = 9; Z = 1.35, p = .180; I² = 78%). Overall subgroup differences based on sidedness were significant (p = .04); no significant difference was seen in left (p = .53) and right-sided (p = .22) superior pRNFL (Figure 2B).

Temporal Peripapillary Retinal Nerve Fibre Thickness (pRNFL)

No difference in the temporal RNFL thickness was found between the AUD group and the control group. (SMD = −0.25; 95% CI = −0.76, 0.27; k = 7; Z = 0.94, p = .35; I² = 91%). Overall subgroup differences based on sidedness were significant (p = .004), but no significant difference in temporal pRNFL thinning was found in either the left side (p = .61) or right side (p = .73) when analysed separately (Figure 2C).

Inferior Peripapillary Retinal Nerve Fibre Thickness (pRNFL)

No difference in the inferior RNFL thickness was found between the AUD group and the control group. (SMD = −0.24; 95% CI = −0.51, 0.02; k = 9; Z = 1.81, p = .07; I² = 70%). Subgroup differences based on sidedness were not significant (p = .14) (Figure 2D).

Nasal Peripapillary Retinal Nerve Fibre Thickness (pRNFL)

Nasal RNFL was found to be significantly thinner in the AUD group than in the control group. (SMD = −0.36; 95% CI = −0.58, −0.13; k = 9; Z = 3.01, p = .002; I² = 56%). Overall subgroup differences based on sidedness were significant (p = .03). Specifically, the left side nasal pRNFL was found to be significantly thinner in the AUD group compared to controls (p < .0001) (Figure 2E).

Average Macular Thickness

Average MT was found to be significantly thinner in the AUD group than in the control group. (SMD = −0.62; 95% CI = −0.95, −0.29; k = 5; Z = 3.66, p = .0002; I² = 50%). Overall subgroup differences based on sidedness were significant (p = .02). Average MT was significantly thinner in the AUD group compared to the control group in both left (p = .0001) and right (p < 0.0001) sides (Figure 3A).

Figure 3.

GCIPL: Ganglion cell inner plexiform layer thickness, RNFL: Retinal nerve fibre layer, Alcohol: Alcohol use disorder group, Control: Healthy control group, SD: Standard deviation, IV: Inverse variance, CI: Confidence interval, Random: Random effects model, df: Degrees of freedom.

Macular Thickness-central

No difference in the central MT was found between the AUD group and the control group. (SMD = −0.24; 95% CI = −0.51, 0.04; k = 5; Z = 1.70, p = .09; I² = 31%). Overall subgroup differences based on sidedness were not significant (p = .17). Right-sided central MT was significantly thinner in the AUD group compared to controls (p = .007) (Figure 3B).

Macular Ganglion Cell-inner Plexiform Layer (mGCIPL)

mGCIPL was found to be significantly thinner in the AUD group than in the control group. (SMD = −0.19; 95% CI = −0.33, −0.06; k = 3; Z = 3.66, p = .004; I² = 67%) (Figure 3C).

Macular Retinal Nerve Fibre Thickness (mRNFL)

No difference in the mRNFL thickness was found between the AUD group and the control group. (SMD = −0.19;95% CI = −0.50, 0.12; k = 3; Z = 1.20, p = .23; I² = 44%) (Figure 3D).

Summary of the Main Effects

Table 1 provides a summary of the results of the meta-analyses carried out, along with the number of studies included, sample size (both patients/controls and eyes), SMD (95% CI), p value and heterogeneity scores.

Publication Bias

For parameters of total pRNFL, inferior pRNFL and MT central, the Doi plots with LFK index suggest low publication bias as LFK index <1, for all other parameters, LFK index >2, suggesting high publication bias (Table S5).

Sensitivity Analysis

Leave-one-out analysis (Table S6) showed that ‘influential’ data points were present for the pooled analysis of the inferior subregion of the pRNFL, central MT and mRNFL. For all these three analyses, leaving one study each caused the p value to change from non-significant to significant.

Meta-regression

Meta-regression was performed to assess the effect of age, amount of alcohol and duration of AUD on the various retinal parameters. None of these parameters showed significant correlation with pRNFL and MT (Table S7).

Certainty of Evidence

Total pRNFL showed moderate CoE. All other parameters were shown to have low CoE (Table S8).

Discussion

A study of retinal morphology in patients with AUD can give us many insights into its pathophysiology and also prognosis. To the best of our knowledge, this is the first published systematic review and meta-analysis that assessed retinal morphology in patients with AUD using OCT. The findings of the current review provide promising evidence for retinal morphological measures, as assessed by OCT, to be biomarkers for AUD. Our meta-analysis revealed significant thinning of the total pRNFL, nasal pRNFL, mGCIPL and average MT in individuals with AUD, compared to HC.

Our review found that significant differences (thinning) were seen in both disc and macular measures. In the general population, the disc or inner retina has been strongly associated with brain white matter (WM) structure and function. In contrast, the macular or outer retina is mainly associated with grey matter (GM).²² Perhaps, lower WM and GM volumes have been consistently reported in patients with AUD; WM loss, typically in frontal areas, is more specific to alcohol compared to other substances.²³ We found that the thinning of pRNFL and macula in patients with AUD reflects the WM and GM abnormalities in them, respectively. Potentially, retinal assessment using OCT, owing to its relatively low cost, ease and feasibility of examination, may be used as a screening measure in the assessment of alcohol related neurodegeneration.

Moreover, our findings revealed that pRNFL and macular thinning in AUD show variable eye-sidedness, on subgroup analysis. The significant differences (thinning) in total and nasal pRNFL were restricted only to the left eyes. Although there was a lower average MT in both right and left eyes, lower central MT was restricted to the right eyes. Therefore, although speculatively, there is a distinct laterality effect: The inner retina (disc) is particularly affected in the left eyes and the outer retina (macula) is more affected in the right eyes. This laterality distinction in AUD has also been noted for the pattern of cognitive dysfunction and impulsivity, which is associated with right hemisphere damage,^24–25 and for craving, which correlates with left hemispheric dysfunction.²⁵ This distinction further aligns with the WM–GM distinction that is found in AUD. Sullivan and Pfefferbaum²⁶ have shown that in AUD patients, frontocerebellar GM structures and frontal superior WM fibre systems are primarily affected. The authors conclude that these changes contribute to accelerated ageing in AUD patients. Thinner retinal layers in AUD patients found in our study also might suggest the same. Intriguingly, AUD is associated with accelerated ‘DNA methylation GrimAge’²⁷ and shortened telomere length,²⁸ both of which are epigenetic predictors of lifespan.^29–30 Perhaps, both these markers are associated with ageing-related neurodegenerative brain changes on magnetic resonance imaging (MRI).^31–32 Specifically, our study showed moderate CoE of the significant thinning in pRNFL, which is an inner retinal layer. Age-related lower brain volume and altered WM network structure have indeed been shown to be associated with thinner inner retinal layers.³³

Cognitive functions were assessed by only one study¹⁶ and hence they were not analysed. Additionally, exploring the correlation between retinal findings and cognitive impairment, as noted in Álvarez-Sesmero et al.¹⁶ could further establish OCT as a biomarker for alcohol related brain damage. Further, we found that there is a significant influence of age, duration of alcohol use and daily alcohol consumption on retinal thinning. As the number of studies needed for meaningful interpretation of meta-regression is not adequate, this finding needs to be explored further in future studies.

Limitations and Strengths

Several limitations of this meta-analysis must be acknowledged. First, the small number of studies with varying outcomes measured in each study may limit the generalisability of the findings. This fact also contributes to the ‘very low’ ratings on the CoE. Second, the definition of ‘case’ is not consistent across studies. Third, the cross-sectional nature of the included studies precludes establishing causality between alcohol use and retinal changes. Longitudinal studies are needed to determine whether these changes progress with continued alcohol consumption and also whether they reverse with abstinence. Fourth, the influence of confounding factors, such as tobacco use or psychiatric comorbidities, was not consistently controlled across studies, potentially affecting the results. Tobacco use itself can cause retinal thinning; the included studies had variable eligibility criteria with respect to tobacco use, thus impacting the results. Fifth, though meta-regression was carried out to estimate the effect of moderators, the analysis is underpowered due to the small number of studies. Final, the exclusion of one study due to insufficient outcome data highlights the need for standardised reporting in OCT studies. The strengths of this review include its rigorous methodology— adherence to PRISMA guidelines, comprehensive assessment of multiple OCT parameters, use of appropriate measures to assess publication bias, sensitivity analysis and assessment of CoE using GRADE-Pro.

Conclusions

This meta-analysis demonstrates that AUD is associated with significant retinal thinning, particularly in the pRNFL and mGCIPL, as detected by OCT. Although promising, these findings should be considered preliminary due to the limitations stated above. Our findings indeed do support the use of OCT as a non-invasive tool to study alcohol related neurodegeneration and potentially monitor disease progression or treatment outcomes. Future research should focus on longitudinal designs, standardised OCT protocols, better reporting of OCT parameters (separately for both eyes, quadrant wise, both macula and disc related parameters, reporting details of all layers), uniform ‘case’ description for AUD, uniform reporting of severity of AUD and well defined exclusion criterion (for other substances, particularly tobacco, ocular diseases, chronic disorders such as diabetes, hypertension) to better elucidate the relationship between alcohol consumption and retinal changes.

Supplemental Material

Supplemental material for this article available online.

Supplemental Material

Supplemental material for this article available online.

Footnotes

Acknowledgements

We thank Dr Lokesh Kumar Singh for providing us access to Scopus and Embase, which helped in completing the search.

Data Sharing Statement

Individual study-level data is available and can be shared on a reasonable request made to the corresponding author.

Declaration of Conflicting Interests

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

Declaration Regarding the Use of Generative AI

None used.

Ethical Approval

Not applicable.

Funding

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

Patient Consent

Not applicable.

Prior Presentations

None.

PROSPERO Registration

CRD42024577392.

Publicly Available Data

None.

Simultaneous Submission to Another Journal or Resource

No.

References

Wolfe

, Menon

, Oto

, . Alcohol and the central nervous system. Pract Neurol, 2023; 23: 273–285. DOI: 10.1136/pn-2023-003817.

Rehm

, Mathers

, Popova

, . Global burden of disease and injury and economic cost attributable to alcohol use and alcohol-use disorders. Lancet, 2009; 373: 2223–2233. DOI: 10.1016/S0140-6736(09)60746-7.

Holloway

, Douglas

, Rafferty

, . Ethanol induces neuroinflammation in a chronic plus binge mouse model of alcohol use disorder via TLR4 and MyD88-dependent signaling. Cells, 2023; 12: 2109. DOI: 10.3390/cells12162109.

Vasilescu

and Macovei

. The perspective of using optical coherence tomography in ophthalmology: Present and future applications. Diagnostics (Basel), 2025; 15(4): 402. DOI: 10.3390/diagnostics15040402.

Sheehan

, Bannai

, Silverstein

, . Neuroretinal alterations in schizophrenia and bipolar disorder: An updated meta-analysis. Schizophr Bull, 2024; 50: 1067–1082. DOI: 10.1093/schbul/sbae102.

Tan

, Schwitzer

, Conart

, . Study of retinal structure and function in patients with major depressive disorder, bipolar disorder or schizophrenia: A review of the literature. J Fr Ophtalmol, 2020; 43: e157–e166. DOI: 10.1016/j.jfo.2020.04.004.

Ahuja

, Kumar

, . Effect of chronic alcohol and tobacco use on retinal nerve fibre layer thickness: A case-control study. BMJ Open Ophthalmol, 2016; 1: e000003. DOI: 10.1136/bmjophth-2016-000003.

Özsoy

and Alim

Optical coherence tomography findings in patients with alcohol use disorder and their relationship with clinical parameters. Cutan Ocul Toxicol, 2020; 39: 54–60. DOI: 10.1080/15569527.2019.1700379.

Abbas

, Hefnawy

and Negida

. Meta-analysis accelerator: A comprehensive tool for statistical data conversion in systematic reviews with meta-analysis. BMC Med Res Methodol, 2024; 24: 243. DOI: 10.1186/s12874-024-02356-6.

10.

Wells

, Shea

, O’Connell

, . The Newcastle-Ottawa Scale (NOS) for assessing the quality of non-randomized studies in meta-analyses. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.html.

11.

Review Manager (RevMan) Version 5.4.1. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2020.

12.

Posit team. RStudio: Integrated Development Environment for R. Boston, MA: Posit Software, PBC, 2023. Retrieved from http://www.posit.co/.

13.

The jamovi project. jamovi. (Version 2.5) [Computer Software]; 2024. Retrieved from https://www.jamovi.org.

14.

GRADEpro GDT: GRADEpro Guideline Development Tool [Software]. Ontario, Canada: McMaster University and Evidence Prime, 2025. Retrieved from https://www.gradepro.org

15.

González-Garcia

, Vilela

, Beltran

MÁ

, . Retinal function assessment in alcohol use disorder patients. J Drug Alcohol Res, 2019; 8: 236070.

16.

Álvarez-Sesmero

, Povedano-Montero

, Arias-Horcajadas

, . Retinal nerve fiber layer in patients with alcohol use disorder. Appl Sci, 2019; 9: 5331. DOI: 10.3390/app9245331.

17.

Orum

and Kalenderoglu

Decreases in retinal nerve fiber layer thickness correlates with cumulative alcohol intake. J Addict Dis, 2020; 38: 400–410. DOI: 10.1080/10550887.2020.1776083.

18.

Liu

, Huang

, Wang

, . The changes in retinal nerve fiber layer and macular thickness in Chinese patients with alcohol dependency. Drug Alcohol Depend, 2021; 229: 109130. DOI: 10.1016/j.drugalcdep.2021.109130.

19.

Şahin

, Karadere

, Yıldız

, . Evaluation of the retinal nerve fiber layer with optic coherence tomography in patients with alcohol use disorder. J Fr Ophtalmol, 2021; 44: 665–669. DOI: 10.1016/j.jfo.2020.11.009.

20.

, Lee

, Fong

AHC

, . The impact of alcohol consumption on retinal nerve fibre layer and macula ganglion cell-inner plexiform layer thickness on optical coherence tomography imaging. Invest Ophthalmol Vis Sci, 2024; 65: PB0095–PB0095.

21.

Stuart

, Luben

, Warwick

, . The association of alcohol consumption with glaucoma and related traits: Findings from the UK biobank. Ophthalmol Glaucoma, 2023; 6: 366–379. DOI: 10.1016/j.ogla.2022.11.008.

22.

Mauschitz

, Lohner

, Koch

, . Retinal layer assessments as potential biomarkers for brain atrophy in the Rhineland Study. Sci Rep, 2022; 12: 2757. DOI: 10.1038/s41598-022-06821-4.

23.

Pando-Naude

, Toxto

, Fernandez-Lozano

, . Gray and white matter morphology in substance use disorders: A neuroimaging systematic review and meta-analysis. Transl Psychiatry, 2021; 11: 29. DOI: 10.1038/s41398-020-01128-2.

24.

Oscar-Berman

and Marinković

. Alcohol: Effects on neurobehavioral functions and the brain. Neuropsychol Rev, 2007; 17: 239–257. DOI: 10.1007/s11065-007-9038-6.

25.

Gordon

. Laterality of brain activation for risk factors of addiction. Curr Drug Abuse Rev, 2016; 9: 1–18. DOI: 10.2174/1874473709666151217121309.

26.

Sullivan

and Pfefferbaum

Alcohol use disorder: Neuroimaging evidence for accelerated aging of brain morphology and hypothesized contribution to age-related dementia. Alcohol, 2023; 107: 44–55. DOI: 10.1016/j.alcohol.2022.06.002.

27.

, Quach

, Wilson

, . DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany, NY), 2019; 11: 303–327. DOI: 10.18632/aging.101684.

28.

Topiwala

, Taschler

, Ebmeier

, . Alcohol consumption and telomere length: Mendelian randomization clarifies alcohol’s effects. Mol Psychiatry, 2022; 27: 4001–4008. DOI: 10.1038/s41380-022-01690-9.

29.

Vaiserman

and Krasnienkov

Telomere length as a marker of biological age: State-of-the-art, open issues, and future perspectives. Front Genet, 2021; 11: 630186. DOI: 10.3389/fgene.2020.630186.

30.

Nannini

, Joyce

, Zheng

, . Alcohol consumption and epigenetic age acceleration in young adults. Aging (Albany, NY), 2023; 15: 371–395. DOI: 10.18632/aging.204467.

31.

Salih

, Galazzo

, Petersen

, . Telomere length is causally connected to brain MRI image-derived phenotypes: A Mendelian randomization study. PLoS One, 2022; 17: e0277344. DOI: 10.1371/journal.pone.0277344.

32.

Sant’Anna Barbosa Ferreira

, van Dongen

, den Braber

, . Epigenetic age acceleration in peripheral blood correlates with brain-MRI age acceleration. Brain, 2025: awaf069. DOI: 10.1093/brain/awaf069.

33.

van der Heide

FCT

, Khawaja

, Berendschot

TTJM

, . Associations of inner retinal layers with risk of incident dementia: An individual participant data analysis of four prospective cohort studies. Alzheimers Dement, 2024; 20: 211–220. DOI: 10.1002/alz.13167.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

0.00 MB

0.58 MB