Intermittent fasting in diabetic retinopathy: microbiome modulation, mechanisms, and clinical insights

Abstract

Intermittent fasting, a nonpharmacological approach, shows promise for controlling diabetic retinopathy but lacks concrete human evidence directly associating intermittent fasting with eye responses. This review highlights preclinical research insights about intermittent fasting and integrates emerging insights on intermittent fasting and its role as a complementary and alternative therapy with conventional diabetic retinopathy therapies. Animal models emphasize that intermittent fasting exhibits multimodal retinal neuroprotective functions and benefits primarily based on enhancing insulin and glucose metabolism, improving retinal mitophagy and mitochondrial functions, and reducing retinal oxidative and inflammation stress. Moreover, alternate-day fasting, 5:2 dietary formulas, and time-restricted eating plans have shown efficacy against acellular capillaries within db/db mice retinas. Simultaneously, intermittent fasting alters retinal gut microbial and bile metabolite responses and thus endorses a gut and eye relationship. These preclinical and emerging research insights thus project intermittent fasting and retinal therapies routed via intermittent fasting and bile acids modulation as retinal supportive and adjuvant therapies.

Introduction

Diabetic retinopathy is a principal microvascular complication of diabetes that is a leading cause of vision loss, representing a sign of progressive metabolic injury of the neurovascular unit (NVU) of the retina [1–3]. Epidemiological surveys report a high and, in some regions, increasing prevalence of diabetic retinopathy, with population-based estimates of approximately 34.6% for any diabetic retinopathy and 7.0% for proliferative diabetic retinopathy (PDR) [2,4]. Diabetic macular edema (DME) can occur at both nonproliferative and proliferative stages and remains a principal cause of vision-threatening outcomes, underscoring the growing burden of PDR and DME [2,4]. Apart from vision, diabetic retinopathy severities keep pace with systemic vascular risk and, more remarkably, actually predict all-cause as well as vascular death, emphasizing the disease as a harbinger of multiorgan metabolic stress [2,5].

Pathophysiology unifies interrelated vascular, neural, and immune mechanisms. Chronic hyperglycemia activates the polyol, advanced glycation end product, protein kinase C, and hexosamine pathways, propelling pericyte loss, endothelial dysfunction, thickening of the basement membrane, and disruption of the retinal barrier; capillary occlusion and hypoxia further upregulate vascular endothelial growth factor (VEGF) [1,6]. Low-grade inflammation, leukostasis, adhesion molecule and inflammatory cytokine expression [e.g. intercellular adhesion molecule-1, vascular cell adhesion molecule-1, selectins, and tumor necrosis factor-α (TNF-α)], and activation of the microglia play a role in endothelial damage and alteration of the blood-retinal barrier. In addition, emerging evidence would propose that this is driven very early on at the retinal NVU and thus contributes to neurovascular dysfunction [1,7]. It is important to consider that neurodegeneration, cell loss, and mitochondrial and reactive oxygen species (ROS) production can potentially precede micro-vasculopathy and thus diabetic retinopathy needs to be reframed as a neurovascular, not purely vascular, disease [1,7].

Although intravitreal anti-VEGF therapy/agents treatment and panretinal photocoagulation changed prognoses for PDR and DME, there remain limitations: injection burden, cost, visual benefit heterogeneity, as well as the high prevalence of VEGF-independent or inflammation-dominant disease in some eyes [1,2]. Therefore, there is increased interest in systemic, lifestyle-driven approaches that decrease upstream causes, hyperglycemia, hypertension, dyslipidemia, chronic inflammation, and metabolic dysfunction associated with adiposity that define diabetic retinopathy risk and progression [2,7]. Observational nutritional literature links high fiber, Mediterranean-style diets, and fish with a lower risk of diabetic retinopathy, whereas caloric excess is linked with progression, supporting the dietary role in ocular complications [8,9].

Dietary interventions vary considerably in design: continuous energy restriction (CER) involves a sustained daily caloric deficit, whereas intermittent fasting cycles between defined periods of fasting and eating. Intermittent fasting encompasses several protocols, including alternate-day fasting (ADF), the 5:2 regimen (2 nonconsecutive fasting days per week), and time-restricted eating (TRE), which limits food intake to a 6–12-h daily window without necessarily reducing total calories. These approaches have shown benefits in glycaemic control, insulin sensitivity, and blood pressure (BP) beyond their comparable weight-loss efficacy to CER [10,11]. Human TRE studies have demonstrated modest benefits in weight loss and BP outcomes, while mechanism-focused reviews discuss the health benefits for glycemia control and entrainment in different populations in humans [12,13]. Beyond classical risk factors, intermittent fasting remaps cellular stress–response programs (autophagy, mitochondrial biogenesis, and ketone signaling) that could mitigate neuroinflammation and oxidative stress, central to diabetic retinopathy [14,15].

Preclinical evidence: intermittent fasting, the gut–retina axis, and bile acid signaling

In diabetic db/db mice, long-term intermittent fasting prevented hallmark diabetic retinopathy lesions, acellular capillaries, and leukocyte infiltration despite no reduction in hemoglobin A1c (HbA1c), coincident with a reshaped gut microbiota, increased mucin-positive goblet cells and villus length, reduced circulating peptidoglycan, and higher plasma tauroursodeoxycholic acid (TUDCA); retinal G protein-coupled bile acid receptor 1 (TGR5/GPBAR1) is expressed, intermittent fasting lowered retinal TNF-α, and pharmacological TGR5 activation (INT-767) prevented diabetic retinopathy [16,17]. Such observations establish an intermittent fasting/bile acid signaling nexus in which intermittent fasting-driven reshaping of the intestinal microbial landscape, coupled with bacterial metabolism of bile acids (Bas) (such as TUDCA), reduces retinal inflammation/microvascular injury in accordance with TGR5-mediated pathways, now placing the gut–retina axis (GRA) into sharper perspective [16,18,19]. The proposed mechanistic links between intermittent fasting-induced gut microbiome remodelling and retinal protection are summarized in Fig. 1.

Fig. 1.

Proposed GRA linking IF-induced microbiome remodelling to protection against DR. DR, diabetic retinopathy; GRA, gut–retina axis; IF, intermittent fasting.

Evidence from animals and cells points to Sirtuin 1 (SIRT1)–liver X receptor (LXR)–ATP-binding cassette transporter A1 (ABCA1)/ATP-binding cassette transporter G1 (ABCG1)-mediated lipid handling and cholesterol efflux as intermittent fasting-responsive nodes: in retinal models, activation of these nodes prevents disruption of retinal homeostasis, while circadian/feeding time analyses reemphasize the importance of the LXR core in the regulation of lipids during fasting regimens [20,21].

Clinical and epidemiological evidence

From clinical syntheses, intermittent fasting is equally effective to CER for weight loss and is one of the widely researched methods along with ADF and the 5:2 diets. A significant reduction in HbA1c and improvements in glycemic parameters were noticed with short-term studies on intermittent fasting among overweight/obese patients with newly diagnosed type 2 diabetes (T2D). However, safety considerations such as the risk for hypoglycemia warrant intermittent fasting under medical supervision [10,11,22,23]. Nutrition-eye studies agree on the beneficial effects of higher quality diets with higher fiber diets and Mediterranean diets with more fatty fish and unsaturated fatty acids in relation to the prevention of diabetic retinopathy, with suggested mechanisms involving the mitigation of retinal oxidative and inflammatory stress, indicating the likely synergy between intermittent fasting timing and nutritional content. The evidence on antioxidant-specific effects remains heterogeneous within systematic syntheses [8,9]. Human data on microbiome–diabetic retinopathy links are still early but increasingly consistent [24]. Cross-sectional studies show a trend toward lower alpha diversity in sight-threatening diabetic retinopathy than in diabetic controls [24]. When participants are grouped by the Bacteroidetes/Firmicutes (B/F) ratio (≈1.05 as a cutoff), the drop in diversity becomes statistically significant, especially within PDR [24]. Results from the longitudinal analyses show that reduced levels of the butyrate-producing bacterial genus Butyricicoccus and lowered plasma acetate at baseline predict an increased 2-year incidence of diabetic retinopathy, indicating a protective role for short-chain fatty acid (SCFA) producing pathways [25].

Mendelian randomization, a method that uses genetic variants as instrumental variables to infer causality while minimizing confounding from observational bias, has identified protective associations between specific gut bacterial taxa and diabetic retinopathy, including Christensenellaceae and Peptococcaceae, lending further support to a causal gut–retina pathobiology [26,27]. Reviews covering broader diabetic complications concur on the following: gut dysbiosis, metabolic endotoxemia (lipopolysaccharide), and impaired intestinal barrier integrity are associated with insulin resistance and vascular inflammation, which could credibly be extended into the retinal microvasculature [18,28].

Mechanistic intersections: ketogenesis, lipid metabolism, and circadian alignment

Circadian biology also has specific implications with regard to the retina because it contains clocks in the retinal photoreceptors, retinal inner cells, and retinal pigment epithelium (RPE). By consolidating feeding into a defined daily window, intermittent fasting can re-align peripheral metabolic rhythms with the light–dark cycle. This re-entrainment may influence retinal perfusion, barrier integrity, and inflammatory tone, providing a plausible mechanistic link between intermittent fasting and diabetic retinopathy-related pathways [12,21,29]. In practical terms, qualitative research has pointed out facilitators and barriers for adherence to TRE in real life, knowledge that is crucial for implementing protocols, while trial synopses have shed light on heterogeneity, the short length of studies, and scarce data in T2D, including the risk for hypoglycemia events [11,30].

At the mechanistic intersection with diabetic retinopathy, intermittent fasting-induced changes in ketogenesis and ketone sensing, enhanced insulin sensitivity, stimulation of autophagy, and nuclear factor kappa-B suppression, described in animal models, correlate with the underlying mechanisms for neuronal resilience and vascular endothelial function [14,31]. There is experimental retinal evidence that activation of the SIRT1–LXR pathway improves cholesterol efflux (ABCA1/ABCG1), decreases inflammation, and protects electrophysiological function (electroretinography/optokinetic nystagmus) [20].

Complementarily, bile acid-based neuroprotection preserves retinal microvasculature and modulates glial responses, supporting actionable molecular bridges from metabolic interventions to retinal protection [3]. Parallel ‘gut-eye’ frameworks catalogue dietary and microbiome-targeted strategies; probiotics, prebiotics, microbial-derived (postbiotic) therapies, bile acid pathway modulation, and fecal microbiota transplantation (FMT) as potential adjuncts to dietary/metabolic interventions aimed at ocular disease [32,33]. Finally, systems-level multiomics show that aging reshapes immune and glycaemic networks in a nonlinear manner with major transitions across midlife, supporting life-course tailoring when deploying intermittent fasting [34,35]. In parallel, circadian-focused syntheses demonstrate that aligning feeding with the light–dark cycle, for example, TRE within less than 12-h daily windows, re-entrains peripheral clocks and synchronizes metabolic programs, motivating explicit time-of-day considerations in intermittent fasting protocols [21,27].

This narrative review addresses the underexplored GRA as a mechanistic link between intermittent fasting and diabetic retinopathy, integrating emerging microbiome and bile acid pathways, including TUDCA, that have been less emphasized in prior reviews.

Considering these complicating factors, we would like to propose intermittent fasting as a potentially viable nutritional therapy for the treatment of diabetic retinopathy, acting on the upstream factors of the disease’s cardiometabolic syndrome, with a specific focus on its gut–retina pathophysiological nexus. By synthesizing preclinical, epidemiological, and clinical evidence, this review aims to build a scientific rationale for testing intermittent fasting as an adjunct therapy in human diabetic retinopathy trials.

Methods

This article is a narrative review synthesizing emerging evidence on intermittent fasting and diabetic retinopathy, with emphasis on gut microbiome-related mechanisms and bile acid signaling (e.g. TUDCA–TGR5). PubMed, Scopus, and Web of Science were searched from inception to December 2025 using combinations of terms related to intermittent fasting (e. g., ‘intermittent fasting’, ‘time-restricted eating/feeding’, ‘alternate-day fasting’, ‘5 : 2’, and ‘fasting-mimicking’) and diabetic retinopathy/retinal outcomes (e. g., ‘diabetic retinopathy’, ‘retina’, ‘blood–retinal barrier’, ‘leukostasis’, and ‘acellular capillaries’), alongside mechanistic terms when relevant (e.g., ‘gut microbiome’, ‘SCFA’, ‘bile acids’, ‘TUDCA’, and ‘TGR5/GPBAR1’).

Studies eligible for inclusion were preclinical (cellular and animal studies) and human trials examining intermittent fasting or fasting-associated methods and publishing retinal outcomes and mechanistic mediators with a biologically plausible link between intermittent fasting and pathways associated with diabetic retinopathy. Non-relevant articles or records without extractable information that could inform the aims of this review were excluded. Literature screening and study identification were conducted by four reviewers, and disagreements were resolved through discussion. Because of heterogeneity in models, interventions, and outcomes, evidence was synthesized qualitatively, and no meta-analysis was performed. A formal risk-of-bias assessment was not undertaken, given the narrative scope of the review.

Intermittent fasting: mechanisms and types

Intermittent fasting regimens alternate periods of substantial energy restriction with intervals of usual intake. Common patterns include: (a) ADF, (b) ‘5:2’ intermittent energy restriction with 2 low-intake days per week, and (c) time-restricted feeding (TRF) in which all calories are consumed within a consistent less than or equal to 12-h window. These have been linked with weight loss, an improvement in insulin sensitivity, as well as an attenuation of inflammation within the system, all of which are relevant to diabetic retinopathy risk reduction [10,36]. ADF regimens commonly include consecutive days of very low-calorie or fasting periods interspersed with ‘eat as much as you like’ days, as outlined within a systematic review and various trials [10]. The 5:2 diet typically involves reducing energy intake to about 500–600 kcal on 2 nonconsecutive days of the week (i.e. approximately 20–30% of daily intake), with ad libitum feeding on the other 5 days [37]. In this review, we use the term TRE to describe human regimens in which daily energy intake is confined to a consistent window of less than or equal to 12 h, and TRF to refer to analogous interventions in animal models. To provide an overview of the main intermittent fasting regimens discussed in this review, their principal mechanisms, and the strength of the supporting evidence, a summary is presented in Table 1.

Table 1.

Summary of intermittent fasting regimens, mechanistic effects, and evidence

Regimen	Description	Key mechanisms	Evidence base	References
ADF	Alternating fasting + ad libitum feeding	Ketogenesis↑, mitochondrial efficiency↑	Robust animal; limited human	[10,14,38]
5:2 diet	2 low-intake days/week (~500–600 kcal)	Metabolic switching, adiposity↓, inflammation↓	Human evidence emerging	[37,39]
TRE (human)	Eating window less than or equal to 12 h/day	Circadian alignment, metabolic flexibility↑	Small trials; heterogeneous	[10,36,40]
TRF (animal)	8–12-h feeding windows in rodents	Clock re-entrainment, F/B ratio↑, barrier integrity↑	Strong preclinical consistency	[16,40]

ADF, alternate-day fasting; F/B, Firmicutes/Bacteroidetes; TRE, time-restricted eating; TRF, time-restricted feeding.

Metabolic and cellular mechanisms

Metabolic switching, which refers to an adaptive change from glucose-centric metabolism to fatty acid- and ketones-centric metabolism with fasting and then normalization, represents a fundamental mechanism associated with intermittent fasting benefits [14,38].

Aligning meal timing with the circadian cycle via TRF (8–12 h feeding windows) re-entrains peripheral clocks, improves metabolic efficiency, and reduces circulating inflammatory markers [10]. On the other hand, TRF is based on restricting energy consumption to a limited time window during the day, typically less than or equal to 12 h (≈8–12 h) in animal models, and helps to resynchronize circadian rhythms and improve metabolic flexibility [40].

In addition, intermittent fasting significantly affects metabolism [14]. The depletion of glycogen levels within white adipose tissue because of fasting enhances the oxidation of fatty acids as well as ketogenesis, hence augmenting mitochondrial biogenesis and cellular resistance to stress [14].

Anti-inflammatory effects

Beyond its metabolic effects, intermittent fasting has also been associated with reductions in systemic inflammatory markers. Across more than 10 preclinical studies, intermittent fasting consistently reduces key inflammatory markers such as TNF-α and interleukin-6 (IL-6) by approximately 20–50% [10,39]. Although these changes are often transient and depend on the specific intermittent fasting regimen, intermittent fasting reprograms the immune response toward an anti-inflammatory phenotype, a redistribution particularly relevant for microvascular complications of diabetes, such as diabetic retinopathy [1].

Gut microbiome and intestinal barrier

There is also evidence for the beneficial role of intermittent fasting regimens in regard to the microbial diversity of the gut [16]. Intermittent fasting has been related to an enrichment of beneficial microbial populations and an improvement in the intestinal barrier in animal studies [16]. For example, in db/db mouse models, intermittent fasting has been shown to increase the Firmicutes/Bacteroidetes ratio, a marker associated with improved metabolic health [16]. Specific changes include boosting the B/F ratio and reducing Bacteroidetes/Verrucomicrobia in db/db mice, coupled with an increase in mucus thickness and goblet cell counts, both markers of an improved intestinal barrier [16]. Contradictory findings continue to appear in human research regarding modifications to Akkermansia muciniphila or Faecalibacterium prausnitzii, or even intestinal barriers [39,41].

Bile acid signaling and the gut–retina axis

Intermittent fasting influences bile acid metabolism, increasing the conjugated secondary bile acid TUDCA and activating its receptor TGR5 in the retina, which leads to reduced TNF-α release and prevention of diabetic retinopathy lesions [16]. Ursodeoxycholic acid has also been implicated in neurovascular protection [3] as well as in the regulation of RPE cell autophagy [16].

TUDCA exerts neurovascular-protective effects, reducing endoplasmic reticulum stress and inducing autophagy in RPE cells [42], as well as in glial and vascular compartments of the retina, thereby preserving capillary structure and preventing degeneration [3].

Translational considerations and human evidence

While animal studies provide consistent evidence of benefit, human trials remain limited, short-term, and heterogeneous [10,36]. Interindividual variability in compliance with fasting protocols and in microbiome responsiveness also complicates the prediction of retinal outcomes [25]. However, emerging evidence indicates that gut microbiota composition may be associated with vision-threatening diabetic retinopathy, for example, decreased abundance of Butyricicoccus and lower plasma acetate levels [25], as well as an increased B/F ratio greater than 1 (≈1.05) in proliferative or clinically significant macular edema diabetic retinopathy [24].

Pathophysiology of diabetic retinopathy and nutritional influences

Core pathophysiological mechanisms

The molecular pathways leading to diabetic retinopathy are closely related to hyperglycemia, as well as the sequelae of oxidative stress, inflammation, and an impaired vasculature, resulting, as a consequence, in the impairment and atrophy of the retinal NVU [1]. The pathophysiological mechanisms leading to diabetic retinopathy are both neuronal and microvascular, representing a complex intersection or network of the pathways that define this disorder [1].

Specifically, the depletion of pericytes has been related to capillary degeneration, thickening of the basement membrane, and dysregulated angiogenesis, resulting in disturbances in the retinal microvasculature with eventual progression to vision impairment [1,43].

Chronic hyperglycemia is cytotoxic to retinal cells and induces oxidative, inflammatory, and metabolic cascades that amplify neurovascular damage in diabetic retinopathy [1,3]. Consequently, the integrity of the microvasculature is impaired, resulting in capillary loss and hypoxia-induced neovascularization [1,9]. In addition, oxidative stress triggers the action of various inflammatory transcription factors, including nuclear factor kappa-B, related to enhanced cytokine production and elevated permeability [1,9,25]. These interconnected mechanisms are summarized in Fig. 2.

Fig. 2.

Core hyperglycemia-driven inflammatory oxidative cascade in DR. Hyperglycemia stimulates the production of ROS and oxidative stress, thereby activating NF-κB and inducing the expression of VEGF and major inflammatory mediators (TNF-α, IL-6, and ICAM-1/VCAM-1). The cumulative effect results in the disruption of the BRB, neovascularization, and the development of acellular capillaries. BRB, blood–retinal barrier; DR, diabetic retinopathy; ICAM-1, intercellular adhesion molecule-1; IL-6, interleukin-6; NF-κB, nuclear factor kappa-B; ROS, reactive oxygen species; TNF-α, tumor necrosis factor-α; VCAM-1, vascular cell adhesion molecule-1; VEGF, vascular endothelial growth factor.

Moreover, oxidative stress is intimately connected with chronic inflammation seen in diabetic retinopathy [1]. Patients with diabetic retinopathy show an increase in IL-6, TNF-α, and IL-1β levels, all pointing to the progression of diabetic retinopathy [1,44]. The mentioned cytokines are involved in glial cell activation, leukostasis, and intraretinal leakage, all leading to retinal injury [1]. There is an interaction between the intraretinal leakage because of inflammation and the leakage because of hyperglycemia, culminating in the formation of acellular capillaries and pathological neovascularization [1].

Nutritional influences and links to intermittent fasting

The development and progression of diabetic retinopathy may, furthermore, be affected by dietary and nutritional components [9]. Diets rich in antioxidants, including polyphenols, vitamin C, vitamin E, and carotenoids, are inversely related to the development of diabetic retinopathy, although evidence related to individual antioxidants has been inconclusive [9]. On the other hand, omega-3 polyunsaturated fatty acids demonstrate anti-inflammatory and antiangiogenic activities against the retinal microvessels, thus preventing retinal microvascular injury [9,43]. Additionally, a Mediterranean dietary regimen, characterized by a high intake of fruits, vegetables, whole grains, and unsaturated fats, has been recognized to reduce the prevalence of diabetic retinopathy [9].

Of particular interest is the potential role of intermittent fasting, a condition now increasingly appreciated as a potential diabetic retinopathy pathophysiological modulator [14]. Intermittent fasting increases the sensitivity of an organism’s tissues to insulin, decreases systemic glycemia, and causes a transition toward ketone bodies, all of which are potential retinal protectants against oxidative and inflammatory injuries [14,45]. Intermittent fasting alters the gut microbial populations and increases levels of TUDCA, a bile acid that activates retinal TGR5 signaling and inhibits TNF-α, thereby acting as a retinal protectant [16]. The latter compound, TUDCA, enhances retinal protectants and autophagy signals against retinal injury [3,42].

Consequently, glial cell activation should be decreased by intermittent fasting, thus maintaining the neurovascular retinal architecture in preclinical model retinal degeneration and diabetic retinopathy [3,42]. Notably, protective nutritional patterns such as the Mediterranean diet and intermittent fasting regimens likely exert synergistic effects through complementary anti-inflammatory and metabolism-improving mechanisms, offering a stronger defense against diabetic retinopathy progression. Although preclinical evidence exists supporting this protective potential, clinical evidence is relatively limited among humans [41]. The duration of treatment, variability in fasting regimens, and compliance are believed to be major hindrances to the successful development of a clinical approach [41,46]. Population-based prospective studies have established an association between dysbiosis and diabetic retinopathy severity, implying a potential therapeutic approach targeted at the GRA by fasting or changing diets, respectively [22,23].

Likewise, when it comes to antioxidant and omega-3 supplement studies, the findings are varied, depending on the dose, type, and study; hence, an integrative approach is advised [8,43]. Together, these findings make a case for the role of dietary and metabolism-based therapies in managing diabetic retinopathy pathogenesis [3]. The additive approach that comes with a combination of intermittent fasting, an antioxidant-rich diet, and an omega-3 supplement may hold promise as a strategy to attack angiogenesis, as well as the other aforementioned components, simultaneously [9]. Nutraceutical strategies focusing on antioxidants, omega-3 fatty acids, the Mediterranean diet, or intermittent fasting may have a potential protective role against retinal damage [14]. A multimodality therapeutic strategy targeting systemic metabolic engineering, the gut microbiome, and retinal protection may work as an adjunct modality to prevent diabetic retinopathy progression [3,9,14,16,47].

Preclinical evidence: intermittent fasting and diabetic retinopathy

Growing preclinical evidence indicates that intermittent fasting may protect against diabetes-associated retinal microvascular dysfunction through several mechanistic axes. A summary of key preclinical studies investigating the effects of intermittent fasting on retinal microvascular outcomes in diabetic models is provided in Table 2.

Table 2.

Summary of key preclinical studies on intermittent fasting and diabetic retinopathy

Study (model)	IF regimen	Key retinal outcomes	Proposed mechanisms
Beli et al., 2018 (db/db mice) [16]	ADF	↓ Acellular capillaries, ↓ leukostasis, and prevention of DR lesions	↑ Plasma TUDCA, activation of retinal TGR5, ↓ retinal TNF-α
Hammer et al., 2021 (db/db mice and retinal endothelial cells) [20]	Fasting-mimicking diet/cycles	↓ Retinal cholesterol accumulation, ↓ inflammatory markers, and alleviated microvascular dysfunction	Activation of SIRT1/LXRα, ↑ ABCA1/ABCG1-mediated cholesterol efflux
Li et al., 2025 (NaIO3-induced retinal degeneration model) [48]	IF (pretreatment)	Mitigated RPE/photoreceptor loss, preserved visual function, attenuated glial activation	↓ ROS-related transcriptional dysregulation, reduced oxidative stress
Huston et al., 2024 (aging mouse model) [49]	TRF	Maintained photoreceptor electrophysiological activity; inconsistent effects on retinal inflammation	Alignment of feeding–fasting cycles with circadian rhythms

ABCA1, ATP-binding cassette transporter A1; ABCG1, ATP-binding cassette transporter A1; ADF, alternate-day fasting; DR, diabetic retinopathy; IF, intermittent fasting; LXRα, liver X receptor α; ROS, reactive oxygen species; SIRT1, Sirtuin 1; TGR5, G protein-coupled bile acid receptor 1; TNF-α, tumor necrosis factor-α; TRF, time-restricted feeding; TUDCA, tauroursodeoxycholic acid.

This table summarizes key preclinical studies exploring the effects of intermittent fasting regimens on retinal outcomes in diabetic models, including the mechanisms involved, such as inflammatory modulation, mitochondrial biogenesis, and changes in the gut microbiome.

Importantly, retinal-specific evidence supporting intermittent fasting’s effects should be attributed to a small set of experimentally verified studies; broader intermittent fasting reviews and circadian/TRE literature can inform systemic mechanisms, but do not substitute for retina-focused preclinical data [16,20] (Fig. 3; Table 3).

Fig. 3.

Proposed gut–retina axis linking IF to retinal protection in DR. IF-related shifts in microbial metabolites and BAs (↑ TUDCA) may activate retinal TGR5 signaling and reduce inflammatory/endothelial activation markers (e.g. ↓ TNF-α). Evidence is mainly preclinical; human retinal endpoints are limited, and some effects may be independent of HbA1c. DR, diabetic retinopathy; HbA1c, hemoglobin A1c; IF, intermittent fasting; TGR5, G protein-coupled bile acid receptor 1; TNF-α, tumor necrosis factor-α; TUDCA, tauroursodeoxycholic acid.

Table 3.

Clinical studies on intermittent fasting and diabetic retinopathy (human trials and observational studies)

No.	Author (s), year	Study type	Participants	IF type/duration	Main outcomes	DR relevance
1	Sutton et al. (2018) [51]	RCTs	8 men with prediabetes (overweight)	Early TRE (6-h eating window, ~08:00–14:00) vs. 12-h control/5 weeks	Improved insulin sensitivity and β-cell function, BP ↓, oxidative stress ↓, and no weight loss	Systemic cardiometabolic improvements that may indirectly reduce DR risk
2	Carter et al. (2018) [56]	RCT	137 adults with T2D	Intermittent energy restriction (2 days/week, 500–600 kcal) vs. CER (1200–1500 kcal/day)/12 months	HbA1c ↓ and fasting glucose ↓ in both groups; intermittent restriction is noninferior to continuous restriction.	Indirect DR benefit via improved long-term glycemic control
3	Trepanowski et al. (2017) [60]	RCT	100 metabolically healthy obese adults (86 women, 14 men)	ADF (25% of energy needs on ‘fast’ days, 125% on ‘feast’ days) vs. daily calorie restriction vs. control/12 months	Similar weight loss to daily calorie restriction; no significant differences in fasting glucose, insulin, or other cardiometabolic markers	Neutral regarding DR; ADF not superior to continuous restriction for metabolic risk factors
4	Cienfuegos et al. (2020) [61]	RCT	58 adults with obesity	4- or 6-h TRE (afternoon/evening eating windows) vs. control/8 weeks	~3% weight loss, energy intake ↓ (~550 kcal/day), insulin resistance ↓ (HOMA-IR), oxidative stress ↓	Metabolic improvements (weight, insulin resistance, oxidative stress) may help delay microvascular complications, including DR
5	Arnason et al. (2017) [62]	Pilot Trial	10 adults with T2D (mostly women; mean BMI in the obese range)	ADF/ 2 weeks	Improved fasting glucose and markers of oxidative stress; short duration, no HbA1c data	Early feasibility for DR prevention
6	Gabel et al. (2019) [63]	Pilot RCT	23 obese adults	8-h TRE (10:00–18:00 eating window)/12 weeks	Weight ↓ (~2–3%), systolic BP ↓; no significant changes in fasting glucose, insulin, HOMA-IR, or lipids	Modest weight and BP improvement may support retinal vascular health; no clear glycemic benefit
7	Parr et al. (2020) [57]	Clinical Trial	24 adults with T2D (19 completed)	9-h TRE (mostly 10:00–19:00)/4 weeks after 2-week baseline	Feasible and safe; on adherent days, energy intake ↓; no significant changes in body weight or HbA1c over 4 weeks	Demonstrates short-term feasibility and safety of TRE in T2D; DR benefit remains unproven
8	Hutchison et al. (2019) [64]	RCT	88 women with overweight/obesity	Intermittent energy restriction (IF70/IF100) vs. CER [continuous daily energy restriction to ~70% of estimated requirements (DR70)]/8-week weight-loss phase (after 2-week lead-in)	Greater reductions in weight, fat mass, total and LDL cholesterol, and NEFA with IF70 vs. DR70; no clear between-group differences in insulin sensitivity	Indirect benefit via weight and lipid improvements; glycemic effects are modest
9	Wilkinson et al. (2020) [12]	Pilot	19 adults with metabolic syndrome	10-h TRE/12 weeks	Weight ↓ (~3%), waist circumference ↓, BP ↓, LDL, and non-HDL cholesterol ↓; trends toward improved fasting glucose and HbA1c (not statistically significant)	Improved cardiometabolic profile and atherogenic lipids; potential adjunct for reducing DR risk, but glycemic and ocular outcomes untested

ADF, alternate-day fasting; BP, blood pressure; CER, continuous energy restriction; DR, diabetic retinopathy; HbA1c, hemoglobin A1c; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; IF, intermittent fasting; LDL, low-density lipoprotein; NEFA, non-esterified fatty acids; RCTs, randomized controlled trials; T2D, type 2 diabetes; TRE, time-restricted eating.

It is crucial to note inconsistencies and limitations across preclinical models. While intermittent fasting prevents diabetic retinopathy lesions independently of HbA1c in some models (e.g. db/db mice), glycemic control remains a significant confounder in others (e.g. Streptozotocin (STZ)-induced models.

In db/db mice, cyclic intermittent fasting treatment elevated circulating TUDCA levels, activated the TGR5 receptor in the retina, localized to the ganglion cell layer, lowered the levels of TNF-α in the retina, and inhibited the lesions of diabetic retinopathy, including acellular capillaries and leukostasis, which were noted to be independent of HbA1c levels, as indicated in Table 3. These retina-specific outcomes align with the right-hand panel of Fig. 3.

Fasting and fasting-mimicking paradigms have also been shown to activate the SIRT1–LXR–ABCA1/ABCG1 lipid-handling axis in retinal and vascular cells, which enhances cholesterol efflux, improves endothelial function, and reduces vascular leakage in experimental models [20] (Fig. 4; Fig. 5).

Fig. 4.

The SIRT1–LXR–ABCA1/ABCG1 axis and endothelial stability in DR. Metabolic signals derived from IF activate SIRT1, which promotes nuclear LXRα/β activity and upregulates ABCA1/ABCG1. This process enhances cholesterol efflux to apolipoprotein A1 (ApoA-I)/HDL. In retinal microvessels, these changes lead to tighter endothelial junctions (↑claudin-5 and occludin), reduced permeability and leukostasis, and decreased ICAM-1/VCAM-1 levels, along with reduced ROS/NF-κB signaling. The evidence primarily comes from preclinical studies. ABCA1, ATP-binding cassette transporter A1; ABCG1, ATP-binding cassette transporter G1; ApoA-1, apolipoprotein A-1; DR, diabetic retinopathy; HDL, high-density lipoprotein; IF, intermittent fasting; ICAM-1, intercellular adhesion molecule-1; LXRα/β, liver X receptor α/β ; NF-κB, nuclear factor kappa-B; ROS, reactive oxygen species; SIRT1, Sirtuin 1; VCAM-1, vascular cell adhesion molecule-1.

Fig. 5.

Indirect (IF → TUDCA → TGR5) and direct TUDCA routes to retinal protection. IF remodels the bile acid pool (↑TUDCA), engaging retinal TGR5 to suppress TNF-α, ICAM-1/VCAM-1, leukostasis, and acellular capillaries while preserving the BRB; effects that may be partly independent of HbA1c (left path). Independently, direct TUDCA delivery to the retina/RPE reduces ER stress, promotes ATG5-dependent autophagy, supports neuroprotection and glial modulation (right path). Both routes converge on preserved retinal microvasculature with ↓ inflammation. ATG5, autophagy-related 5; BRB, blood–retinal barrier; ER, endoplasmic reticulum; IF, intermittent fasting; HbA1c, hemoglobin A1c; ICAM-1, intercellular adhesion molecule-1; RPE, retinal pigment epithelium; TGR5, G protein-coupled bile acid receptor 1; TNF-α, tumor necrosis factor-α; TUDCA, tauroursodeoxycholic acid; VCAM-1, vascular cell adhesion molecule-1.

TUDCA itself, independently of intermittent fasting treatment, shows neurovascular-protective effects in various models of retinal disease and induces autophagy in RPE cells [3,42]. In describing bile acid signaling pathways, it thus appears that two aspects should be distinguished (see Figs. 3 and 5): aspect (a) intermittent fasting changed bile acid composition (elevated TUDCA), and subsequent TGR5 receptor activation (Table 3) [16], and aspect (b) the direct treatment effect of TUDCA on various models of the retina described in refs [3,42].

Additional preclinical research involving the retina further supports the neuroprotective and anti-inflammatory roles of fasting/mimetics. TRF has been shown to maintain photoreceptor-mediated electrophysiological activity in an aging mouse model, but it did not have a consistent effect on the levels of retinal inflammation in that particular model [49]. In a NaIO₃-induced model of retinal injury, intermittent fasting reduced glial cell activation and ROS production, indicating neuroprotection in that particular model. With respect to retinal biomarkers, Beli et al. reported reductions in pro-inflammatory cytokines, notably TNF-α, in association with intermittent fasting, consistent with the anti-inflammatory profile summarized in Table 3 [50].

The limitations of the preclinical literature should also be clarified and include: animal model similarities/differences (db/db and STZ-induced diabetic models), variations of the intermittent fasting protocol (a.k.a. ADF and 2-day cycles), structural/functional outcomes assessed, and the absence of human retinal data, which together pose challenges for translation and must be acknowledged when interpreting these findings (as outlined in Table 3).

In brief, retina-targeting preclinical evidence supports three converging, mechanism-driven pillars of intermittent fasting-related retinal protection:

1. modulation of bile acid signaling via TGR5, with inhibition of TNF-α and reductions in acellular capillaries, as listed in Table 3 [16];

2. activation of SIRT1/LXR/ABCA1/ABCG1 pathways, improving lipid metabolism and reinforcing endothelial integrity, as discussed in Table 3 [20];

3. enhancement of autophagy-mediated neuroprotection in the RPE and neural retina, supported by data on TUDCA treatment, as summarized in Table 3 [3,42].

Beyond retina-specific models, broader intermittent fasting literature reports consistent systemic benefits on weight, insulin sensitivity, BP, and inflammatory markers [27,51–53], with epigenetic remodeling of metabolic tissues [54] and circadian alignment of feeding–fasting cycles [55] providing additional mechanistic routes relevant to microvascular protection. Clinical trial evidence in T2D and obesity is summarized in Table 3 [56–64].

More recent work has begun to extend these concepts both back to the retina and forward to broader aging and circadian biology. In a sodium iodate–induced model of oxidative stress-driven outer retinal degeneration that mimics features of age-related macular degeneration rather than diabetic retinopathy, Li et al. [48] showed that pretreatment with an intermittent fasting regimen mitigated loss of RPE and photoreceptors, preserved visual function, and attenuated microglial and Müller-cell activation in association with reduced ROS-related transcriptional dysregulation. This retina-focused experimental evidence dovetails with circadian and dietary-restriction frameworks such as those synthesized by Takahashi and colleagues [65], who emphasize that the beneficial effects of energy restriction on health span and cardiometabolic aging are critically modulated by clock-controlled nutrient-sensing pathways. Together, these studies support the view that intermittent fasting/TRE paradigms interface with oxidative stress, inflammatory, and circadian axes that are highly relevant to retinal microvascular and neurodegenerative pathology, even though direct diabetic retinopathy models remain sparse.

On the clinical side, intermittent energy restriction and TRE regimens have been tested primarily in individuals with T2D, obesity, or insulin resistance rather than with retinal endpoints. In patients with T2D, a 12-month randomized noninferiority trial by Carter et al. found that a 2 days/week intermittent energy restriction protocol produced HbA1c reductions and weight loss comparable to those achieved with continuous daily energy restriction [56], while a shorter feasibility study by Parr et al. [57] demonstrated that a 9-h time-restricted-eating window was implementable in this population, with reduced energy intake on adherent days and modest overall changes in glycemic control.

Narrative and systematic syntheses by Varady et al. [58] and Patterson and Sears [59] indicate that ADF, 5:2 regimens, and TRE generally induce mild-to-moderate weight loss, improve insulin resistance and several cardiometabolic risk factors, and are usually safe when appropriately monitored. Randomized controlled trials in adults with overweight or obesity, including a year-long comparison of ADF versus daily calorie restriction [60], an 8-week trial of 4-h and 6-h TRF [61], a pilot study of daily intermittent fasting in T2D with improvements in body weight and at-goal fasting glucose [62], and a secondary analysis in insulin-resistant individuals showing greater reductions in fasting insulin and The homeostasis model assessment of insulin resistance with ADF than with continuous restriction [63] are complemented by shorter-term time-restricted-feeding interventions such as those reported by Hutchison and colleagues [64], which point to improvements in glucose tolerance and insulin sensitivity in at-risk individuals.

Emerging human studies and clinical relevance

While preclinical evidence is compelling, human data directly linking intermittent fasting to diabetic retinopathy outcomes remain extremely limited, representing a critical translational gap. The available human data have primarily determined the beneficial effect of these approaches on upstream cardiometabolic risk factors (e.g. glycemia, lipids, BP, and inflammation) that indirectly contribute to diabetic retinopathy pathogenesis [10,12,13,22].

A narrative synthesis of over 15 RCTs in individuals with T2D or obesity shows consistent benefits of various intermittent fasting regimens (e.g. TRE, ADF, 5 : 2) on weight loss, HbA1c reduction, and improvements in BP and lipid profiles. These systemic cardiometabolic improvements are strongly associated with reduced diabetic retinopathy risk and progression. However, it must be emphasized that none of these trials included direct ocular endpoints (e.g. retinal thickness, microaneurysm count), and thus, the direct efficacy of intermittent fasting for diabetic retinopathy in humans remains unproven. Table 4 summarizes selected key RCTs providing this indirect evidence [10,12,13,22].

Table 4.

Selected randomized controlled trials of intermittent fasting in type 2 diabetes (indirect evidence for diabetic retinopathy)

Study and design	Participants (N)	IF regimen and duration	Key metabolic outcomes
Carter et al., 2018 (RCT) [56]	137 adults with T2D	IER (2 days/week, 500–600 kcal) vs. CER (1200–1500 kcal/day)/12 months	HbA1c ↓ comparable between groups (IER noninferior)
Sutton et al., 2018 (RCT) [51]	8 men with prediabetes	Early TRE (6-h window) vs. 12-h control/ 5 weeks	Insulin sensitivity ↑, BP ↓, oxidative stress ↓
Cienfuegos et al., 2020 (RCT) [61]	58 adults with obesity	4- or 6-h TRE vs. control/ 8 weeks	Weight loss (~3%), insulin resistance (HOMA-IR) ↓
Guo et al., 2024 (RCT – EARLY trial) [22]	Adults with early T2D and overweight/obesity	5:2 IF with meal replacement/16 weeks	Significant HbA1c lowering vs. metformin/empagliflozin

BP, blood pressure; CER, continuous energy restriction; HbA1c, hemoglobin A1c; HOMA-IR, homeostasis model assessment of insulin resistance; IER, intermittent energy restriction; IF, intermittent fasting; RCT, randomized controlled trial; T2D, type 2 diabetes; TRE, time-restricted eating.

Various clinical studies have revealed that diverse intermittent fasting schedules, including TRE, ADF, and the 5 : 2 diets, are able to notably lower the fasting glucose levels, HbA1c levels, and triglycerides of patients with T2D [51,53,56]. These findings indirectly validate the presumption that intermittent fasting-induced improvements of metabolism and vasculature can help alleviate the progression of diabetic retinopathy through increased sensitivity to insulin and reduction in oxidative stress. Furthermore, novel evidence suggests that intermittent fasting may increase the abundance of beneficial taxa (e. g., Akkermansia muciniphila) and reduce circulating inflammatory cytokines (e. g., IL-6 and TNF-α), changes that could favor microvascular repair in the diabetic retina [24,25,57]. The underlying mechanisms through which intermittent fasting benefits diabetic retinopathy, including metabolic regulation, reduction of oxidative stress, and modulation of the gut microbiome, are further detailed in Table 5.

Table 5.

Mechanisms of intermittent fasting in diabetic retinopathy

Mechanism	Description	References
Metabolic regulation and insulin sensitivity	Improves insulin sensitivity and glycaemic control and reduces systemic oxidative stress and inflammatory markers, changes that may indirectly lessen microvascular injury in the retina.	[14,16,17,38,51,56]
Activation of autophagy and mitochondrial function	Activates autophagy and adaptive stress–response pathways, improves mitochondrial health, and promotes survival of retinal pigment epithelial and neural retinal cells in experimental models.	[14,20,42]
Impact on gut and bile acid signaling	Modifies the GM and bile acid pattern, enhances the neuroprotective bile acid TUDCA, and provides TGR5-mediated anti-inflammatory and neurovascular protection that is also relevant to the retina.	[16,18]
Gut microbiome and its link to retina	The GRA and its modulation: the effect of disturbances of GM homeostasis and IF-induced changes of the microbiota profile and the production of bacterial metabolites on retinal inflammation.	[19,24,25]
Lipid metabolism and endothelial reinforcement	Stimulates the SIRT1–LXRα–ABCA1/ABCG1 pathway to improve cholesterol efflux, return retinal lipids to normal, and lower endothelial inflammation and microvascular injury.	[20]

ABCA1, ATP-binding cassette transporter A1; ABCG1, ATP-binding cassette transporter G1; GM, gut microbiota; GRA, gut–retina axis; IF, intermittent fasting; LXRα, liver X receptor α; SIRT1, Sirtuin 1; TGR5, G protein-coupled bile acid receptor 1; TUDCA, tauroursodeoxycholic acid.

Limitations and Translational Challenges: The translation between preclinical and human studies is still challenging. Intermittent fasting paradigms are very different between animal and human studies regarding duration and caloric background [13]. There are limitations related to animal species, sex, and age [13]. Pharmacologic TUDCA supplements differ from the microbiome-derived bile acid environment generated by human intermittent fasting, limiting direct comparability [22]. Surrogate endpoints for human studies are mainly focused on recognized metabolic markers and not directly measuring retinal parameters like fundus autofluorescence (FAF)/optical coherence tomography angiography (OCTA) and the thickness of choroids. However, safety issues related to hypoglycemic states of intermittent fasting remain poorly evaluated and may impact adherence and anti-VEGF and Insulin therapy [12]. Thus, future studies are required to follow unified methods and longer-lasting intermittent fasting to provide a concrete connection between fasting and microbiome modulation in animal and human studies for human diabetic retinopathy development. Table 5 summarizes the principal mechanistic pathways linking intermittent fasting to retinal protection in diabetic retinopathy.

The role of gut microbiome in intermittent fasting and diabetic retinopathy

Diabetic retinopathy is one of the frequent microvascular complications in patients with T2D [19,32]. While management of blood glucose levels is important, evidence has been emerging that it is not enough to slow diabetic retinopathy. New studies point to the gut microbiome as a central player in the modulation of systemic inflammation and retinal health, leading to the indication of an intestinal–retina axis that hypothesizes bidirectional interplay between intestinal microbial profiles and retinal physiology [16,19,25]. A schematic overview of the proposed gut–retina axis linking intermittent fasting to retinal protection is provided in Fig. 3. This axis affects retinal health via microbial metabolites, immunomodulation, and modulation of gut barrier integrity, and provides an alternative paradigm to the conventional glucose-centric therapeutic approaches [19]. Intermittent fasting has been termed a nutritional approach that affects gut microbiota and systemic metabolism [16]. In animal research using db/db mice as an example of Type 2 DM, ADF and ad libitum (ad-lib) feeding increased life span and attenuated diabetic retinopathy markers without being affected by the level of HbA1c concentration [16]. More importantly, in these studies, intermittent fasting decreased the penetration of leukocytes as well as the number of acellular (nonnucleated) capillaries within the retina. Together, these changes suggest decreased microvascular inflammation and damage. These alterations are of particular importance under diabetic conditions, when elevated gut permeability may enhance systemic inflammation and retinal microvascular injury [19,25].

More recently, Mendelian randomization studies have sought to demonstrate the existence of possible causal links between specific gut microbes, in particular those with SCFA production capability, and the risk of diabetic retinopathy, thus supporting the conceptual basis for a GRA [24].

Lower abundance of genera belonging to the Bacteroidetes and Verrucomicrobia can also be functionally relevant. While some reports associate Bacteroidetes shifts with carbohydrate handling, signals linking dysbiosis to barrier dysfunction are suggested in reviews [19]. Verrucomicrobia (mainly Akkermansia) is, in general, beneficial, but the decrease during intermittent fasting does not seem to be enough to override the positive effects of increased SCFA-producing Firmicutes [16]. Generally, the modifications in the host microbial communities triggered by intermittent fasting will favor an environment rich in SCFA-producing taxa, such as Butyricicoccus and Faecalibacterium, which are known to promote epithelial integrity and resistance to systemic inflammation and potentially provide an environment protective for the retina [16,23,41].

Intermittent fasting-triggered reshaping of the microbiota modifies metabolic pathways associated with retinal protection. These Firmicute-derived SCFAs provide energy precursors for colonocytes, fortify tight junctions of cells in the intestinal lining, and have anti-inflammatory actions at a systemic level [16]. Furthermore, intermittent fasting controls BA metabolism by increasing TUDCA, a neuroprotective member of the BAs family that binds to TGR5 in the ganglion cell layer, activating downstream anti-inflammatory signaling [16,66].

Upon TGR5 activation, retinal inflammation and vascular damage were lessened regardless of glucose levels, thus providing evidence of a direct mechanism linking gut microbiota alteration with the health of the retina. In addition, intermittent fasting-stimulated microbiome remodeling strengthens systemic immune regulation. Increased SCFA induces the expansion of regulatory T-cells in gut-associated lymphoid tissue, and this is accountable for decreased systemic pro-inflammatory cytokines such as TNF-α, IL-1β, and IL-6 [16].

The development of a favorable microorganism ecosystem and metabolic homeostasis is also in line with the morphological changes in the intestine (such as villi hypertrophy and increased mucosal area) [16]. Human studies support the importance of these microbiota shifts. SCFA-producing genera, e. g., Butyricicoccus and Ruminococcus torques, were lower in diabetic retinopathy subjects than controls, and plasma SCFA concentrations, including acetate, were reduced [25]. The temporal assessment has confirmed the potential of the above microorganisms to predict the incidence of diabetic retinopathy [25].

Longitudinal analysis indicated that these microbial risk markers were predictive of diabetic retinopathy, with a possible translational potential for targeting SCFA-producing bacteria for the prevention or postponement of diabetic retinopathy development [19,25]. These findings argue in favor of the role of the microbiome as a predictive and interventional tool clinically. TUDCA has been recognized as an important metabolite involved in retinal protection. TUDCA suppresses the retinal stress response and maintains microvasculature/neuroglia in preclinical models [3]; it also induces autophagy in RPE cells [42].

The roles of intermittent fasting include the induction of TUDCA expression and triggering TGR5 in the retina. These actions decrease caspase-mediated apoptotic cell death, endoplasmic reticulum stress, and the expression of pro-inflammatory cytokines in diabetic retinopathy and choroidal neovascularization models [66]. TUDCA also increased autophagy in an Atg5-dependent fashion in retinal pigment epithelial cells and reduced oxidative stress in an Atg5-dependent fashion in these cells [42]. These findings suggest that the role of intermittent fasting may include an increase in the creation of a positive microenvironment around the retina. Longitudinal studies linking microbiota composition and metabolite levels to diabetic retinopathy outcomes are needed to develop predictive intervention models [25,32].

Proposed therapeutic targets include prebiotics, probiotics, postbiotics, bile acid pathway modulation, and FMT as adjuncts to dietary management [67–69]. Potential synergy may emerge from integrating intermittent fasting with other interventions, not only with ideal blood glucose control but also with pharmacological advances.

FMT has shown promise in restoring gut bacterial balance and improving insulin sensitivity in diabetes models, though evidence for direct diabetic retinopathy benefit remains insufficient [68]. Postbiotics, including butyrate, similarly demonstrate anti-inflammatory and neuroprotective properties relevant to retinal protection [69].

A cross-sectional study based on the National Health and Nutrition Examination Survey (NHANES), 2009–2018, was conducted to investigate the association of a gut microbiota supportive dietary index with diabetic retinopathy [70]. Subjects who had a diet high in fiber and plant-based foods, both of which are favorable for gut bacteria, had significantly lower likelihoods of diabetic retinopathy, suggesting that dietary intervention to improve gut microbiota can contribute to stabilizing the progression of diabetic retinopathy [70]. Based on this evidence, healthy dietary habits, microbiota-targeted therapies, and regular ophthalmic care could combine to slow disease progression and improve visual outcomes [68].

Challenges and considerations for implementing intermittent fasting in diabetic retinopathy management

Intermittent fasting has attracted considerable attention as a potential strategy to modulate diabetic retinopathy by enhancing systemic metabolism and reducing inflammation [71]. However, its practical application has some challenges and issues related to safety matters [16,19].

A primary concern is hypoglycemia, particularly in individuals treated with insulin or other glucose-lowering medications. This concern is especially critical for insulin-treated individuals, because fasting with unchanged medication schedules can precipitate severe hypoglycemia [41].

Careful patient selection for intermittent fasting is essential. Clinicians should consider usual glycemic control, comorbidities, dietary habits, lifestyle factors, and current medications. Ensuring patient safety through individualized treatment plans is of prime importance.

The compliance of intermittent fasting schemes may be undermined by various cultural and social factors, such as meal times with the family and participation in societal events [16,41]. The education of patients and families concerning the mechanism of fasting, the detection of hypoglycemic symptoms, and the appropriate modification of fasting practices is a very important task that needs to be accomplished in order to convince patients to strictly follow the scheme and prevent the development of adverse events [16]. If patients follow intermittent fasting, the role of professionals with knowledge of behavior and motivation is important because psychological and practical factors play a significant role in compliance [41].

Intermittent fasting may be safely integrated into diabetic retinopathy care provided that it is implemented under close medical supervision, individualized treatment plans, appropriate patient education, and culturally sensitive counseling [71]. Although preliminary studies suggest that intermittent fasting can reduce systemic inflammation and may confer retinal protection, further research is required to clarify its long-term safety, define optimal fasting regimens, and characterize interactions with existing diabetic retinopathy therapies [16,19]. Overall, the main potential benefits and risks of implementing intermittent fasting in the context of diabetic retinopathy care are summarized in Table 6.

Table 6.

Potential benefits and risks of intermittent fasting in the management of diabetic retinopathy

Benefits	Risks
Systemic inflammation reduction	Hypoglycemia (especially with insulin)
Improved insulin sensitivity	Low adherence because of social/psychological barriers
SCFAs reduction/TUDCA increase	Potential nutrient deficiencies (rare)
Possible retinal protection	Requires supervision and medication adjustment

SCFAs, short-chain fatty acids; TUDCA, tauroursodeoxycholic acid.

Comprehensive patient education regarding physiological responses to fasting is essential. Educational programs should ensure that patients recognize the clinical signs and symptoms of hypoglycemia and understand appropriate corrective actions. Education should also include information on the safe adjustment and interruption of fasting programs if required [16].

Importantly, these are also applicable to psychologists and behavior specialists because the aspects of motivation and management of issues influence compliance and overall outcomes [16]. Cultural sensitivity and the response and changes that occur within a standard protocol are also important. Intermittent fasting for the treatment of diabetic retinopathy needs a very high level of supervision and further research into its longer-term safety, fasting practices, and effects [16,19].

Conclusion

At present, while preclinical data and stage I human studies offer evidence supporting that intermittent fasting is a promising nonpharmacological adjunct therapy in diabetic retinopathy treatment, preclinical models suggest that intermittent fasting can improve multiple key pathophysiologic features of diabetic retinopathy, including vascular permeability, neovascularization, and inflammation. Presumably, the protective effects are because of the following: increased sensitivity to insulin, reduced inflammation and oxidative stress, increased autophagy, and the following mechanisms involving the microbiome: production of SCFA and TUDCA/TGR5 signaling. Early studies in humans have also shown that intermittent fasting could positively influence the following metabolism parameters: glycemia, HbA1c, and triglycerides. This is a positive factor, as it is well known that all of these variables predict lower diabetic retinopathy incidence, although the intermittent fasting impact directly related to ocular variables such as retinal thickness or visual acuity has not been fully assessed in a human study.

When considered in relation to other dietary approaches, such as Mediterranean diets or CER diets, intermittent fasting offers its own set of unique benefits. Although the Mediterranean diet and CER diets have been well proven in terms of their efficacy in enhancing overall cardiometabolic profiles as well as managing diabetes, intermittent fasting is an alternative therapeutic strategy that is likely easier to administer, especially when considered in terms of approaches such as TRE or ADF diets. On the one hand, its metabolic and anti-inflammatory responses, as well as its ability to act in a beneficial manner upon the GRA, make intermittent fasting an attractive option to include in a diabetic retinopathy therapeutic strategy.

Intermittent fasting’s integration within current treatments of diabetic retinopathy could be a supplement to current treatments in enhancing glycemic control, improving insulin sensitivity, as well as decreasing oxidative stress, all of which are paramount in the progression of diabetic retinopathy. Moreover, intermittent fasting might also augment current treatments involving anti-VEGF agents as well as other treatments of ocular diseases because of its ability to boost general vascular health. Nonetheless, substantial knowledge gaps remain regarding the use of intermittent fasting alongside current diabetic retinopathy treatments. A large amount of existing knowledge has emerged from short-term trials with small numbers of participants, and there has been a failure to emphasize the measurement of ocular endpoints adequately.

This still needs to be a rigorously implemented randomized controlled trial within the 6–12-month window for assessment of various ocular endpoints (such as OCTA-defined microvascular density, foveal avascular zone, and retinal thickness) concurrently with various -omic analyses and inclusion of a structured hypoglycemia component, especially with exogenous insulin use.

Integrating preclinical and early human evidence positions intermittent fasting as a promising nonpharmacologic nutritional strategy that may help mitigate diabetic retinopathy. Mechanistically, intermittent fasting can improve glycemic control and insulin sensitivity, reduce oxidative stress and inflammation, and activate catabolic stress–response pathways (autophagy, mitochondrial remodeling) while modulating gut–retina communication via bile acid signaling (e.g. TUDCA–TGR5) and SIRT1–LXR–ABCA1/ABCG1–mediated cholesterol efflux. However, evidence in humans remains limited and largely indirect, and DR-specific RCTs with standardized intermittent fasting protocols, safety monitoring (especially hypoglycemia in T2D), and ocular endpoints (e.g. OCTA and ERG) are urgently needed. Intermittent fasting’s promise lies in its low-cost, systemic approach, but awaits validation in DR-specific RCTs. Until such data are available, intermittent fasting should be considered only as an individualized, closely monitored adjunct rather than a standard therapy for diabetic retinopathy.

Acknowledgements

O.A, D.G., A.H, S.M.M., A.S., P.S, E.G., G.Z., and S.A.: investigation and writing the manuscript. M.S., A.M., Z.J., and P.D.: supervision. S.J.K.: conseptualization, edition, and supervision.

Conflicts of interest

There are no conflicts of interest.