Introduction
The olfactory system has a unique capacity for regeneration, which is dependent on the maintenance and differentiation of the olfactory neuroepithelium.Reference Paschaki, Cammas, Muta, Matsuoka, Mak and Rataj-Baniowska1 Vitamin A and its metabolites, known as retinoids, are critical for this process, supporting epithelial integrity and neurogenesis. Pre-clinical research has established this biological link; retinoid deficiency leads to irreversible damage to the olfactory epithelium in animal models, while supplementation appears to promote functional recovery following injury.Reference Asson-Batres, Zeng, Savchenko, Aderoju and McKanna2
This biological plausibility has spurred interest in exploring the clinical utility of vitamin A and its analogues as a potential therapeutic option for patients with olfactory dysfunction. Despite this, the clinical evidence base is fragmented and contains conflicting results from studies varying widely in design and quality.
Therefore, this systematic review was conducted to comprehensively synthesise the existing clinical evidence to determine the efficacy and safety of vitamin A and its analogues in treating olfactory dysfunction. This work aims to provide a robust summary of the current literature, identify key evidence gaps, and inform future research directions in this important area of rhinology.
Materials and methods
Protocol and registration
The review protocol was developed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and registered with the International Prospective Register of Systematic Reviews (PROSPERO) [ID: CRD420251017758]. The flowchart is outlined in Figure 1.
PRISMA flowchart.
Figure 1 Long description
The process of identifying new studies through databases and registers is outlined in a flowchart. It is divided into three main sections: Identification, Screening and Included. 1. Identification: - Records identified from databases (n equals 28) and registers (n equals 0). - Records removed before screening due to duplicate records (n equals 10), records marked as ineligible by automated tools (n equals 0) and records removed for other reasons (n equals 0). 2. Screening: - Records screened (n equals 18). - Records excluded (n equals 0). - Reports sought for retrieval (n equals 18). - Reports not retrieved (n equals 0). - Reports assessed for eligibility (n equals 18). - Reports excluded (n equals 10). 3. Included: - New studies included in review (n equals 8). - Reports of new included studies (n equals 0).
Eligibility criteria
Studies were eligible if they evaluated the efficacy of vitamin A (or vitamin A analogues) in adults with olfactory dysfunction of any cause. Eligible study designs included randomised controlled trials (RCTs), controlled clinical trials, prospective cohort studies, and case series with 10 or more participants. Included studies were required to report measurable outcomes of olfactory function using validated tools (e.g. Sniffin’ Sticks, University of Pennsylvania Smell Identification Test (UPSIT)) or clearly defined subjective improvement. Animal studies, reviews, editorials and studies lacking extractable outcome data were excluded.
Search strategy
A systematic literature search was performed in PubMed, Embase, Web of Science, and the Cochrane Library from inception up to September 2025. The search strategy included terms related to ‘vitamin A’, ‘retinoids’, ‘anosmia’, ‘hyposmia’ and ‘olfactory dysfunction’. No language or date restrictions were applied. Reference lists of included articles were also screened for additional eligible studies. The full database-specific search strategy is provided as online supplementary material.
Study selection
Two reviewers (KSJ and DP) independently screened titles and abstracts of identified records. Full texts of potentially eligible studies were retrieved and assessed against the pre-defined inclusion and exclusion criteria. Disagreements were resolved by discussion and, where necessary, consultation with a third reviewer (HM).
Data extraction
Data from included studies were extracted independently by two reviewers (KSJ and DP) using a standardised, pre-piloted data extraction form. This was used to collect relevant information from each study, including study characteristics, participant demographics, intervention details and outcome data. Corresponding authors were contacted for clarification where needed.
Quality assessment
The methodological quality and risk of bias of included studies were assessed independently by two reviewers (KSJ and DP) using the Cochrane Risk of Bias tool (RoB 2) for RCTs and the Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) Assessment for observational studies. Disagreements were resolved by discussion.
Results
Study selection and characteristics
A total of 28 records were identified through database searches. After removal of 10 duplicates and screening of titles and abstracts, 18 full-text articles were assessed for eligibility. Eight studies met the inclusion criteria and were included in the systematic review.Reference Duncan and Briggs3–Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 The 8 included studies were published between 1962 and 2024 and enrolled a total of 589 participants.Reference Duncan and Briggs3–Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 A PRISMA flowchart is shown in Figure 1, with the included studies summarised in Table 1.
Summary of included studies; AE = adverse event; BTT= butanol threshold test; IM = intra-muscular; IU = international units; OT = olfactory training; RCT = randomised controlled trial; SIT = Smell Identification Test; TDI = Threshold, Discrimination, Identification; UPSIT = University of Pennsylvania Smell Identification Test
Table 1 Long description
Columns list study and year, design with aetiology, sample size, intervention versus control, and main findings for vitamin A or isotretinoin and smell outcomes. Duncan et al. 1962 case series of 64 participants, 56 treated, used intra muscular vitamin A 100000 international units weekly for 6 weeks with or without oral dosing, reporting 89 percent improved, best with intra muscular dosing, poor response in congenital or traumatic anosmia, and no serious adverse events. Reden et al. 2012 double blind randomised controlled trial with 52 participants compared oral vitamin A 10000 international units daily for 3 months with placebo and reported no significant difference in Threshold Discrimination Identification change or in gaining 6 points, and no effect by aetiology. Hummel et al. 2017 retrospective cohort with 170 participants compared intra nasal vitamin A 10000 international units daily for 8 weeks plus olfactory training with olfactory training alone, reporting in post infectious cases 37 percent improved versus 23 percent with a p value of 0.03, and no benefit in the post traumatic group. Kartal et al. 2017 prospective observational study of 33 acne participants without baseline loss used oral isotretinoin 0.5 to 0.8 milligrams per kilogram per day for 3 months and reported increased Sniffin Sticks scores with a p value less than 0.001, a trend toward less hyposmia, and common adverse event cheilitis. Taheri et al. 2024 double blind randomised controlled trial in post covid 19 with 90 participants in three groups compared olfactory training plus placebo, olfactory training plus oral vitamin A 10000 international units daily, and placebo only for 3 months, reporting improved normosmia rates with olfactory training with or without vitamin A and no added benefit of vitamin A. De Sousa et al. 2023 open label pilot study of 47 persistent post covid participants compared olfactory training alone, olfactory training plus corticosteroids, olfactory training plus corticosteroids plus multivitamin, and olfactory training plus intra nasal vitamin A 15000 international units plus vitamin E for 3 months, reporting all improved and greater gain for olfactory training plus vitamin A plus vitamin E versus olfactory training alone with a p value of 0.03 and no adverse events. Chung et al. 2023 open label pilot randomised controlled trial in long covid with 26 participants, 22 analysed, compared oral vitamin A 25000 international units daily for 14 days plus aerosolised olfactory training for 4 weeks with olfactory training and with observation, reporting vitamin A plus olfactory training better than olfactory training or control for butanol threshold test and smell identification test, functional magnetic resonance imaging increased olfactory connectivity, and no adverse events. Kondo et al. 2024 cross sectional acne study with 70 participants compared oral isotretinoin 40 milligrams daily for 3 months with no treatment and reported lower University of Pennsylvania Smell Identification Test scores in the isotretinoin group, hyposmia more prevalent, and no anosmia.
Efficacy of vitamin A on olfactory function
The included studies demonstrated mixed evidence for the efficacy of vitamin A and its analogues in treating olfactory dysfunction.Reference Duncan and Briggs3–Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 Five studies reported improvement in olfactory performance following vitamin A or retinoid therapy, particularly in post-infectious aetiologies and when combined with olfactory training.Reference Duncan and Briggs3–Reference Kartal, Yaşar, Kartal, Özcan and Borlu7 However, two RCTs found no statistically significant added benefit over control interventions,Reference Reden, Lill, Zahnert, Haehner and Hummel9, Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 and one cross-sectional study suggested a potential negative association between systemic isotretinoin exposure and olfactory function.Reference Kondo, Miot, Frare, Garcia, Yamakami and Fornazieri8
Early evidence from Duncan et al. (1962) reported improvement in 89 per cent of patients treated with intra-muscular vitamin A, with superior outcomes in post-viral cases compared to congenital or traumatic causes.Reference Duncan and Briggs3 In a large retrospective cohort, Hummel et al. (2017) found that intra-nasal vitamin A combined with olfactory training led to a significantly greater proportion of responders (TDI gain 5.5 points or greater) in post-infectious patients compared to training alone (37 per cent vs 23 per cent, p = 0.03).Reference Hummel, Whitcroft, Rueter and Haehner4 De Sousa et al. (2023) similarly reported superior olfactory perceptual threshold gains when intra-nasal vitamins A and E were combined with olfactory training (mean change 4.4 vs 2.9, p = 0.03).Reference De Sousa, Machado, da Costa JC, Silva, Pinto and Coutinho5
Chung et al. (2023) conducted a three-arm pilot study on long COVID-related anosmia. Patients receiving high-dose oral vitamin A (25 000 IU/day for 14 days) with olfactory training showed significantly greater improvements in butanol threshold test scores compared to both olfactory training alone (mean difference = 3.2, p = 0.009) and untreated controls (mean difference = 4.4, p < 0.001). The combination group also demonstrated a significant within-group improvement in the Smell Identification Test (p = 0.009). Neuroimaging revealed increased functional connectivity in olfactory networks, and spectroscopy showed elevated N-acetylaspartate to creatine ratios in the orbitofrontal cortex, correlating with olfactory gains and significantly differing between groups (p = 0.021). These findings support a potential central and peripheral regenerative effect.Reference Chung, Zhang, Wong, Sridhar, Lee and Leung6
Outcomes were more variable for systemic retinoids. Kartal et al. (2017) reported significant improvement in TDI scores after three months of oral isotretinoin use in a cohort of acne patients (p < 0.001).Reference Kartal, Yaşar, Kartal, Özcan and Borlu7 In contrast, Kondo et al. (2024) observed significantly lower UPSIT scores and a higher prevalence of hyposmia in isotretinoin users compared to controls (p < 0.001), raising concerns about potential dose-dependent olfactory suppression.Reference Kondo, Miot, Frare, Garcia, Yamakami and Fornazieri8
Two RCTs did not demonstrate a clear benefit from vitamin A supplementation. Reden et al. (2012) found no statistically significant difference in changes to TDI scores between vitamin A and placebo groups (p = 0.47), despite reports of subjective improvement in both arms.Reference Reden, Lill, Zahnert, Haehner and Hummel9 Taheri et al. (2024) evaluated vitamin A combined with olfactory training in patients with COVID-19-related anosmia and reported no significant between-group difference in objective outcomes, although both groups showed within-group improvements over time.Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10
Overall, five studies support some degree of benefit from vitamin A or its analogues, especially in post-infectious settings and when used adjunctively.Reference Duncan and Briggs3–Reference Kartal, Yaşar, Kartal, Özcan and Borlu7 However, two RCTs failed to show superiority over placebo or training alone,Reference Reden, Lill, Zahnert, Haehner and Hummel9, Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 and one study indicated possible harm.Reference Kondo, Miot, Frare, Garcia, Yamakami and Fornazieri8
Dosage, duration and route of administration
The included studies exhibited considerable heterogeneity in the dosage, duration and route of administration of vitamin A and its analogues. Both oral and topical (intra-nasal) routes were employed, with varying durations of treatment and specific retinoid compounds.
Oral administration of vitamin A or its analogues was reported in five studies. Reden et al. (2012) administered oral vitamin A at a dose of 10 000 IU/day for 3 months.Reference Reden, Lill, Zahnert, Haehner and Hummel9 Taheri et al. (2024) used oral vitamin A at 10 000 IU/day, also for 3 months, in combination with olfactory training.Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 Chung et al. (2023) provided oral vitamin A 25 000 IU/day for a shorter duration of 14 days, combined with aerosolised diffuser olfactory training for 4 weeks.Reference Chung, Zhang, Wong, Sridhar, Lee and Leung6 Two studies investigated oral isotretinoin: Kartal et al. (2017) used systemic isotretinoin at 0.5–0.8 mg/kg/day for 3 months, while Kondo et al. (2024) examined patients receiving oral isotretinoin at 40 mg daily in a cross-sectional design, with treatment duration varying among participants.Reference Kartal, Yaşar, Kartal, Özcan and Borlu7, Reference Kondo, Miot, Frare, Garcia, Yamakami and Fornazieri8
Intra-nasal administration of vitamin A was reported in two studies. Hummel et al. (2017) combined olfactory training with intra-nasal vitamin A at 10 000 IU/day for 8 weeks.Reference Hummel, Whitcroft, Rueter and Haehner4 De Sousa et al. (2023) explored tailored therapy that included olfactory training plus intra-nasal vitamin A and E for three months.Reference De Sousa, Machado, da Costa JC, Silva, Pinto and Coutinho5
Safety and adverse effects
The assessment of safety and adverse effects associated with vitamin A and its analogues for olfactory dysfunction was limited by inconsistent and incomplete reporting across the included studies.
Of the eight studies, four (Reden et al., 2012; Taheri et al., 2024; Hummel et al., 2017; Duncan and Briggs, 1962) provided no explicit information regarding adverse events or stated that none were reported.Reference Duncan and Briggs3, Reference Hummel, Whitcroft, Rueter and Haehner4, Reference Reden, Lill, Zahnert, Haehner and Hummel9, Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 Two studies (Chung et al., 2023; de Sousa et al., 2023) explicitly stated that no adverse events were observed or reported during or after the interventions.Reference De Sousa, Machado, da Costa JC, Silva, Pinto and Coutinho5, Reference Chung, Zhang, Wong, Sridhar, Lee and Leung6
The study by Kondo et al. (2024), which investigated oral isotretinoin in acne patients, did not report specific adverse events within its study participants but did acknowledge that known side effects of isotretinoin include nasal dryness and rhinitis, implying these could be mild to moderate.Reference Kondo, Miot, Frare, Garcia, Yamakami and Fornazieri8
Kartal et al. (2017), also investigating isotretinoin, was the only study to provide detailed reporting on adverse events. In this cohort of 33 patients receiving isotretinoin, up to 12 participants (36.4 per cent) experienced mucocutaneous side effects, including dryness, cheilitis and conjunctivitis. Cheilitis was noted as the most common adverse event. These adverse events were generally classified as mild to moderate, although they did lead to study dropout in some cases.Reference Kartal, Yaşar, Kartal, Özcan and Borlu7
Overall, the available evidence regarding the safety and tolerability of vitamin A and its analogues in the context of olfactory dysfunction is scarce and lacks systematic reporting. While isotretinoin, as a potent vitamin A analogue, is known to induce mucocutaneous side effects, the specific safety profile of vitamin A supplementation for olfactory indications remains largely underreported in the current literature. Future research should prioritise rigorous and comprehensive reporting of adverse events to establish a clearer safety profile.
Quality assessment overview
The quality assessment of the eight included studies revealed a heterogeneous mix of study designs with varying risks of bias (Table 2). Among the three randomised studies, Reden et al. (2012) demonstrated the lowest risk of bias (‘low to some concerns’) due to its double-blind, placebo-controlled design, while Taheri et al. (2024) and Chung et al. (2023) raised ‘some concerns’ owing to issues including dropout, open-label design, uneven allocation and baseline differences.Reference Chung, Zhang, Wong, Sridhar, Lee and Leung6, Reference Reden, Lill, Zahnert, Haehner and Hummel9, Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 The two negative RCTs are clinically important, but their sample sizes, attrition and phenotype heterogeneity mean that imprecision and Type II error cannot be excluded.Reference Reden, Lill, Zahnert, Haehner and Hummel9, Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 The observational studies generally presented a higher risk of bias; four studies, including prospective and retrospective cohorts and a case series, were rated as ‘critical’ primarily due to the absence of control groups, non-randomised treatment allocation and the use of subjective outcome measures. Although the cross-sectional study by Kondo et al. (2024) used a matched control group and validated measures, its design inherently limits causal inference.Reference Kondo, Miot, Frare, Garcia, Yamakami and Fornazieri8 Overall, the methodological quality was mixed, with significant limitations identified in the majority of non-randomised studies.
Quality assessment of included studies; RCT = randomised controlled trial
Table 2 Long description
Quality assessment is reported by study year, study design, overall risk of bias, and a brief rationale. Reden et al. 2012, an RCT, is rated low to some concerns, citing double blind randomised placebo controlled design, validated outcome measure, standardised protocol, and limited dropout information. Taheri et al. 2024, an RCT, is rated some concerns, citing double blind randomised controlled design, validated outcome measures, significant dropouts, and baseline severity differences between groups, with a note that the negative result may be underpowered for modest subgroup effects. Chung et al. 2023, an RCT, is rated some concerns, citing open label pilot design, small sample size, high dropout in the standard care group, uneven allocation, and acknowledged baseline differences, with a note that findings should be considered hypothesis generating. Kartal et al. 2017, a prospective cohort, is rated critical due to absence of a control group. Hummel et al. 2017, a retrospective cohort, is rated critical due to retrospective design and lack of true control group or randomisation. Kondo et al. 2024, a cross sectional study, is rated low to some concerns, citing matched control group and validated objective measures, and noting that cross sectional design allows association but not causal inference. De Sousa et al. 2023, a prospective cohort, is rated critical, citing standardised testing and clear outcome reporting, and noting non randomised tailored intervention allocation based on endoscopic findings and patient preference. Duncan et al. 1962, a case series, is rated critical, citing non standardised treatment, lack of control group, and subjective outcome measures.
Discussion
The findings of this review provide important insights into the current state of evidence for using vitamin A and its analogues to manage olfactory dysfunction. While pre-clinical evidence suggests a role for retinoids in olfactory epithelial regeneration, the available clinical evidence does not support the routine clinical use of vitamin A or retinoid analogues for the improvement of olfactory function.
A recurring theme in our findings is the potential for vitamin A to enhance outcomes in specific patient subgroups, particularly in post-infectious aetiologies where regenerative capacity is likely preserved. These findings align with the proposed biological mechanism of retinoids in promoting olfactory epithelial regeneration by stimulating basal progenitor cell differentiation and enhancing sensory neuron turnover.Reference Paschaki, Cammas, Muta, Matsuoka, Mak and Rataj-Baniowska1
Consistent improvements were observed in studies targeting post-infectious aetiologies, especially when vitamin A was used as an adjunct to olfactory training, where regenerative capacity is likely to be preserved.Reference Duncan and Briggs3–Reference De Sousa, Machado, da Costa JC, Silva, Pinto and Coutinho5 The neuroimaging and neurochemical data from Chung et al. (2023) further support this hypothesis, suggesting that vitamin A may amplify the neurorehabilitative effects of sensory stimulation by promoting both peripheral and central neural plasticity.Reference Chung, Zhang, Wong, Sridhar, Lee and Leung6
The findings from Kondo et al. (2024) raise further concerns about possible olfactory suppression with long-term or high-dose systemic retinoid use, highlighting the need to distinguish between therapeutic and potentially toxic exposures.Reference Kondo, Miot, Frare, Garcia, Yamakami and Fornazieri8 The assessment of safety and tolerability is hampered by inconsistent and often minimal reporting of adverse events. While isotretinoin is known to cause mucocutaneous side effects, the safety profile of vitamin A supplementation for olfactory indications is largely -reported.
For the practising clinician, these findings reinforce that olfactory training remains the primary evidence-based recommendation for patients with post-infectious olfactory loss.Reference Hummel, Whitcroft, Rueter and Haehner4 While this review found insufficient evidence to support the routine use of vitamin A, the data suggest a potential adjunctive role in select post-infectious cases. Such use should be guided by a shared decision-making process with the patient, acknowledging the current lack of robust supporting evidence and the need for further high-quality trials.
The two negative RCTs should not be interpreted as definitive proof of absence of effect. Reden et al. randomised 52 participants across mixed post-infectious and post-traumatic aetiologies, and Taheri et al. randomised 90 participants across 3 arms with dropouts and baseline imbalance.Reference Reden, Lill, Zahnert, Haehner and Hummel9, Reference Taheri, Naderi, Jonaidi Jafari, Emadi Koochak, Saberi Esfeedvajani and Abolghasemi10 In both cases, treatment effects may have been diluted by heterogeneous phenotypes, spontaneous recovery, concurrent olfactory training and insufficient power to detect modest subgroup effects. Accordingly, the rationale for further RCTs is not that previous trials were uninformative, but that they were underpowered and heterogeneous for determining whether any route-, dose- or phenotype-specific effect exists.
There are several critical areas for future research to address the current evidence gaps regarding vitamin A and its analogues for olfactory dysfunction.
Future trials should be designed around a pre-specified minimum clinically important difference and powered accordingly, rather than simply testing vitamin A in small, heterogeneous groups. Trial design should separate post-infectious, post-traumatic and idiopathic phenotypes; distinguish oral from intra-nasal administration; and evaluate vitamin A, both as monotherapy and as an adjunct to established therapies, such as olfactory training. A notable example is the APOLLO trial, a double-blinded, randomised proof-of-concept study investigating intra-nasal vitamin A for post-viral olfactory loss.Reference Kumaresan, Bengtsson, Sami, Clark, Hummel and Boardman11 This trial, with its robust design and inclusion of objective imaging outcomes such as olfactory bulb volume, provides a model for the kind of higher-quality research needed in this area.
Secondly, future research must standardise the reporting of interventions and outcomes. Studies should clearly specify the type of retinoid used, its formulation, exact dosage and route of administration. Furthermore, the use of validated and widely accepted objective olfactory tests, such as the Sniffin’ Sticks test or the UPSIT, should be mandatory. Reporting of continuous data on change scores (e.g. mean change and standard deviation from baseline) is essential to enable future quantitative meta-analyses.
Thirdly, future research should focus on identifying specific patient populations that might benefit from these interventions. Subgroup analyses in large trials could explore the efficacy of vitamin A in different aetiologies of olfactory dysfunction (e.g. post-viral, post-traumatic, idiopathic) and determine if treatment effects are more pronounced in specific groups.
Finally, and perhaps most importantly, there is an urgent need for comprehensive and consistent reporting of adverse events. Future trials must include a dedicated section for safety reporting, detailing all adverse events, their severity and their relationship to the study intervention. This will allow for a more robust assessment of the true risk-benefit profile of vitamin A and retinoids in this therapeutic context.
By addressing these key areas, future research can provide the high-quality evidence necessary to determine whether vitamin A or retinoid analogues have a role in defined subgroups of patients with olfactory dysfunction.
Conclusion
Based on the available evidence, vitamin A and its analogues cannot be recommended for the routine treatment of olfactory dysfunction. While some studies suggest a potential synergistic effect with olfactory training in post-infectious patients, the evidence is weak and inconsistent. Existing negative RCTs are important but limited by sample size, heterogeneity and imprecision; therefore, further trials should be adequately powered, phenotype-specific and use standardised interventions, validated outcome measures and systematic adverse-event reporting.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0022215126105507.
Authorship contribution
KSJ and DP: equal contribution to conception and design of the work, collecting data, analysis, writing manuscript; HM: design of the work, editing and final approval of the manuscript. All co-authors take full responsibility for all aspects of the study and the final manuscript.
