Age-related macular degeneration (AMD) is one of the leading causes of irreversible vision impairment and loss among older adults (1). Vision impairment and loss caused by AMD imposes a huge economic burden and severely reduces the quality of life (2). Current cutting-edge research confirms the involvement of multiple factors, such as complement, lipids, angiogenesis, inflammation, extracellular matrix remodeling, and metabolomics, in the pathogenesis of AMD (3–9). Thyroid hormone (TH) is an important endocrine messenger for normal development and physiological function. In the retina, TH signaling plays a critical role in regulating optic protein expression in cone cells. In animal models of AMD, THs have been shown to play an important role in the development of the disease, and anti-TH drugs protect the retinal pigment epithelium (RPE) and photoreceptors from damage and apoptosis (9, 10). Therefore, an in-depth study of the relationship between TH signaling and the pathogenesis of AMD and is of great importance for understanding disease pathogenesis and identifying new therapeutic measures. In this study, we provide a narrative review of the literature, existing shortcomings, and future research directions on the relationship between TH and AMD from three dimensions: basic experiment, genetics, and clinical studies.

This narrative review employed a structured but non-systematic approach to literature identification and synthesis. While not adhering to the complete PRISMA guidelines for systematic reviews, we aimed to minimize selection bias through the following steps: We searched PubMed, Web of Science, and Embase databases for articles published between January 1990 and March 2024 using combinations of the terms “thyroid hormone,” “thyroxine,” “TSH,” “FT4,” “thyroid disease,” and “age-related macular degeneration” or “AMD.” Only studies published in English were considered due to practical constraints in translation and consistency in terminology interpretation. Articles were screened based on relevance to the research theme. Eligible evidence included basic experimental studies, genetic analyses, epidemiological cohorts, case–control studies, and clinical investigations directly examining the relationship between thyroid function, thyroid-related medications, and AMD. The exclusion criteria included non-English publications, conference abstracts without full text, and studies not directly addressing the TH–AMD relationship. Reference lists of relevant reviews and included studies were also screened to identify additional publications.

Mendelian randomization (MR) is an emerging method of epidemiological research in recent years, both nationally and internationally, to study the exact causal relationship between a disease exposure factor and the disease, with a focus on investigating causal relationships between risk factors and disease outcomes from a genetic perspective (16). Compared with traditional observations and analyses, the use of MR is less susceptible to confounding or reverse causation and supports causal inference in a genetic sense (17–19).

Li et al. (20) included 33,526 individuals (16,144 in the case group and 17,832 controls) from a genome-wide association study (GWAS) of 72,167 Europeans selected for free thyroxine (FT₄) and TSH-related polymorphisms, with pooled single-nucleotide data on AMD as well as from the GWAS. The results revealed that elevated FT₄ levels were significantly associated with an increased overall risk of AMD (OR = 1.189, 95% CI = 1.058, 1.334, p = 0.005); there was no evidence of a direct causal effect of TSH levels on AMD risk (OR = 0.955, 95% CI = 0.810, 1.125, p = 0.582). The above finding suggests that the FT₄ gene level is positively associated with AMD risk, while there is a lack of clear evidence between TSH and AMD risk. This study establishes, for the first time, a causal relationship between FT₄ and AMD from a genetic point of view, which is more reliable compared to correlation studies, but this study did not confirm the causal relationship between TSH and AMD. In addition, population studies have revealed that rs943080 in the VEGF-A gene is associated with AMD, and these clinical studies reveal that susceptibility genes may be found in the AMD inflammatory and oxidative stress pathways (21).

KG et al. first demonstrated that human RPE cells express TH receptors and speculated that they might be a direct target of TH (22). Recent studies by Tsai et al. (23) have come to a similar conclusion that the two are associated and are emerging as an additional risk factor in the assessment of retinal degenerative diseases (10, 24, 25). Furthermore, a meta-analysis including seven studies (n = 61,993) confirmed that thyroid dysfunction was associated with increased AMD risk (pooled OR ≈ 1.19, 95% CI 1.06–1.33; I² = 80.1%, p < 0.001), although between-study heterogeneity remained high (26). In contrast, the Blue Mountains Eye Study (n ≈ 3,600, 10-year follow-up) did not confirm a consistent association between thyroid function and AMD after multivariable adjustment, particularly for TSH within the reference range (26). Taken together, these findings suggest a stronger and more consistent FT₄–AMD relationship in the Rotterdam cohort, whereas the Blue Mountains data highlight possible population or methodological differences. The inconsistent associations observed between the Rotterdam Study and the Blue Mountains Eye Study may reflect differences in population characteristics and study design. Variations in dietary iodine intake, age distribution, ethnic composition, baseline thyroid status, and duration of follow-up could influence circulating thyroid hormone levels and retinal vulnerability. In addition, differences in outcome definitions and multivariable adjustment strategies may further contribute to the observed heterogeneity (Table 1).

Both clinical and subclinical thyroid dysfunction are common in the general population. In Europe, the collective prevalence of these conditions reaches a level of 3.82%. However, the mechanisms linking thyroid disorders to AMD are unclear. TH are potent regulators of mitochondrial biogenesis and oxidative phosphorylation. Elevated TH signaling can increase mitochondrial respiration and ATP production, thereby enhancing ROS generation. In RPE cells, which exhibit high metabolic demand and limited antioxidant capacity, this shift may overwhelm protective enzymes such as superoxide dismutase and glutathione peroxidase, resulting in oxidative damage and cellular dysfunction. This mechanism provides a plausible biological link between elevated TH activity and increased susceptibility to AMD.

Retinal tissue exhibits susceptibility to oxidative stress, with such damage being a well-established constituent of AMD’s multifaceted pathogenesis. Furthermore, population studies indicate that certain VEGF-A genes are linked to AMD and hold relevance in association studies focusing on TSH ranges. A possible explanation may be the discovery of sensitivity genes in inflammatory and oxidative stress pathways. Additional studies have confirmed that thyroid dysfunction is associated with vascular function, lipids, and atherosclerosis (27) and is a susceptibility factor for the development of AMD (28).

In addition, several clinical studies have explored the perspective of thyroid medications and the occurrence of AMD (10, 25). Thyroxine medications also appear to play a role in influencing the risk of developing AMD, with a population-based study in the Augsburg region reporting a significantly higher risk of other ocular disorders in TH-treated male patients, but the relationship is unclear at this time and needs to be validated by numerous later prospective studies and large samples of data. A separate meta-analysis of 118,420 participants examining thyroid hormone replacement therapy also indicated an association with AMD, but with moderate heterogeneity (I² = 69.0%) (25). In contrast, a study of a large population-based sample of adults aged 50 years showed a correlation between hypothyroidism and AMD, particularly among white women (29). Future studies with larger sample sizes are needed.

Recent therapeutic advances have highlighted the role of nuclear receptors in regulating aging-related pathways, including oxidative stress, inflammation, lipid metabolism, and angiogenesis, many of which are implicated in AMD pathogenesis (30). However, accumulating experimental evidence indicates that excessive or systemic TH signaling exacerbates oxidative damage and mitochondrial dysfunction in RPE and photoreceptors, suggesting that global activation of TH pathways is unlikely to be beneficial (30).

Importantly, the therapeutic relevance of TH receptors should not be interpreted as supporting systemic receptor agonism. Instead, future strategies may involve selective, tissue-specific, or context-dependent modulation of nuclear receptor signaling, aiming to preserve physiological homeostatic functions while avoiding the detrimental effects associated with excessive TH activity (30). At present, there is insufficient evidence to support TH receptor agonists as a therapeutic option for AMD, and local inhibition or fine-tuned regulation of TH signaling appears more consistent with existing experimental findings.

Furthermore, AMD is a complex, multicellular disease involving not only RPE cells but also photoreceptors, choroidal endothelial cells, resident immune cells, and microglia. A comprehensive understanding of nuclear receptor signaling across these cellular compartments will be essential before any receptor-based interventions can be translated into clinical practice (23).

In conclusion, this study describes the correlation between TH and AMD in recent years and new ideas of treatment mechanisms from the perspectives of basic experiments, clinical research, genetic method research, etc. However, due to the complexity of the mechanism of AMD disease itself, coupled with the fact that the current relevant research on the mechanism of hormone-induced AMD doctrine is still in the stage of discovery of the correlation, residual confounding factors cannot be excluded, and the underlying mechanisms of this association remain speculative—whether through direct retinal TH signaling, systemic oxidative stress, vascular alterations, or shared genetic susceptibility, it is not yet clear. Consequently, the lack of TH axis-mediated ocular AMD occurs in the stage of research or remains in the stage of extrapolation. It is believed that, in the future, with the continuous improvement of exploring the specific mechanisms of the two, breakthroughs will be made in the precise treatment of dry AMD.

Some studies have adopted telomere length (TL) and age-related phenotypes as indicators of senescence. As a biomarker of cellular aging, TL is also recognized as a risk factor for age-related diseases. Given that telomere length shortens with advancing age, it has been extensively explored in relation to both cellular senescence and the risk of developing age-related diseases.

The current review includes studies that inevitably have several limitations: First, unmeasured factors such as diet, systemic inflammation, or comorbidities may influence thyroid status and AMD risk despite multivariate adjustment. Second, a majority of studies did not distinguish between early and late AMD or between dry and neovascular AMD, which may have weakened the association. Third, a majority of cohorts assessed FT₄/TSH once at baseline, which may not capture lifetime thyroid function or fluctuations. Fourth, a majority of genetic and cohort data are from European populations; there is limited and sometimes discrepant evidence for Asian and other ancestries. In addition, the observed associations with thyroid medications may reflect, in part, the severity or comorbidity of the underlying thyroid disease rather than the drug effect itself. Finally, formal quality assessment tools (e.g., Newcastle–Ottawa Scale, ROBINS-I) were not applied. Future studies should prioritize multi-ethnic MR analyses and large prospective cohorts in non-European populations, particularly in Asian and iodine-variable regions, to evaluate the generalizability of the observed associations and to explore potential ancestry-specific or environmental modifiers of thyroid hormone–AMD relationships.

Studies have shown that thyroid disease accelerates the onset of age-related phenotypes, including shortened TL, premature onset of senile cataracts, and AMD, and further experimental studies could be conducted in the future to elucidate the mechanisms by which thyroid disease promotes aging, which could help to slow the progression of aging and disease (31). It is also worth noting that internists and endocrinologists often miss ophthalmological treatment before referring patients with thyroid disease to ophthalmologists, and this delay in referral may potentially lead to prolonged treatment of thyroid eye disease. It is inferred that its optimal treatment may also be missed in the case of AMD, which needs to be studied in the future (32). Based on previous studies, we found that genetically predicted higher FT₄ (within the normal range) was associated with later AMD, and we found that this association was predominantly evident in women. However, these may be serendipitous findings, and high normal FT₄ levels may increase the risk of AMD later in life, especially in women, and further research is needed to investigate the role of FT₄ in AMD and potential gender differences (33).