Abstract

Background

Previous studies have reported associations between individual nutrients or specific foods and the risk of age-related macular degeneration (AMD). However, the relationship between overall dietary quality, specifically the alternative Mediterranean diet (aMED) score, and AMD remains unclear. Therefore, this study aimed to investigate the association between the aMED score, as an indicator of overall diet quality, and AMD in the Korean population.

Methods

We conducted a cross-sectional analysis using nationally representative samples of older adults aged鈥夆墺鈥�65 years (895 men and 1,191 women) from the Korea National Health and Nutrition Examination Survey (2017鈥�2018). Food intake and the aMED score were estimated using 24-h recall. AMD was diagnosed by an ophthalmologist based on fundus photography. The associations of aMED score tertiles with AMD were determined using odds ratios (ORs) from multivariate logistic regressions.

Results

Multiple logistic regression analysis revealed a significantly negative association between the aMED score and AMD (adjusted ORs鈥�=鈥�0.58; 95% confidence interval鈥�=鈥�0.39鈥�0.88; p-trend鈥�=鈥�0.021) in older men after adjusting for confounding factors such as age, body mass index, family monthly income, current smoking, alcohol consumption, physical activity, chronic disease status, and energy intake. Notably, this association was exclusively observed in men, and no significant association was observed between the aMED score and AMD in women (adjusted OR鈥�=鈥�0.88; 95% CI鈥�=鈥�0.61鈥�1.29; p-迟谤别苍诲鈥�=鈥�0.691).

Conclusions

This study鈥檚 findings suggest that a high aMED score may be associated with a reduced risk of AMD in older men. Future studies with larger sample sizes and a prospective or interventional design are required to enhance current understanding regarding the association between diet quality and AMD.

Background

Age-related macular degeneration (AMD) inflicts damage upon the macula, leading to severe vision loss in older adults [1, 2]. The prevalence of AMD is anticipated to rise in tandem with the aging global population, projecting to reach 288听million cases by 2040, with a notable risk surge occurring at 50鈥�60 years of age across diverse ethnic groups [3, 4]. This demographic shift underscores the imperative for researchers to address the considerable medical expenses associated with AMD treatment and the resulting impact on affected individuals鈥� quality of life [5, 6].

AMD-associated aging involves the focal deposition of acellular detritus between the retinal pigment epithelium and Bruch鈥檚 membrane [7]. These deposits, namely, drusen, are small yellow accumulations in the macula and peripheral retina. The size, number, and degrees of confluence of drusen are significant risk factors for the development of AMD [8].

Several studies have identified AMD as a consequence of the complex interaction of genetic, environmental, and behavioral factors, among others. Its risk factors include age [9], body mass index (BMI) [10], sex [11], genetics [12], ethnic group [13], hypertension [14], cancer [15], cardiovascular disease [16], diabetes mellitus [17], smoking [18], alcohol [19], physical activity [20], and diet [21].

Previous studies, especially the Age-Related Eye Disease Study (AREDS) and AREDS2, have focused on the relationships between nutritional interventions and AMD [22, 23]. Antioxidant nutrients (e.g., carotenoids such as lutein and zeaxanthin, fruits, vegetables, vitamins C and E, zinc, and fatty acids such as n-3 polyunsaturated fatty acids) have been shown to reduce AMD risk [24,25,26,27]. Research has recently emphasized the broader impact of overall dietary patterns on AMD risk, with various dietary patterns, such as the Oriental and Western patterns, and diet quality demonstrating reduced odds of AMD [28,29,30].

The Mediterranean diet, characterized by the abundant consumption of plant foods (vegetables and fruits), legumes, whole grains, nuts, and olive oil as a fat source; low-to-moderate intake of fish, poultry, dairy products, and alcohol; and low intake of red meat [31], is associated with decreased oxidative stress [32]. Systematic analyses have reported that adherence to a Mediterranean diet is associated with a decreased risk of AMD progression [33].

While most studies on the relationship between the Mediterranean diet and AMD have been conducted in Western regions, such as the United States and Europe [34, 35], few prominent studies have investigated the association of dietary patterns with AMD in Asian populations. One such study by Kang & Kim [30] analyzed the relationship between specific dietary components, including fish and legumes, and AMD based on the Korean population [30]. This study focused on individual food components and found higher fish and legume consumption to be associated with reduced odds of early and late AMD, respectively [30].

In Korea, the prevalence of AMD in adults aged鈥�>鈥�40 years is 13.19%, indicating that more than one in 10 people have the disease. Since age itself is the predominant risk factor, AMD prevalence tends to increase with age, yielding prevalence rates of 3.6%, 11.3%, 20.3%, and 31.4% in individuals in their 40s, 50s, 60s, and 70s, respectively [36]. Considering the rapid increase in the 65-year-old population in Korea, investigating the relationship between the Mediterranean diet score and AMD in the older population is imperative. Therefore, this study aimed to analyze diet quality using the alternative Mediterranean diet (aMED) score, which assesses overall dietary quality by consolidating multiple food group scores into one aggregate score, is adapted to the Korean diet [37], and examines the association between this score and AMD in adults aged鈥�>鈥�65 years.

Methods

Study design and participants

This study used data from the Korea National Health and Nutrition Examination Survey (KNHANES, 2017鈥�2018) [38]. The KNHANES, a nationally representative cross-sectional survey comprising health interviews, health examinations, and nutrition surveys, employs a stratified, multi-stage clustered probability design to ensure representation of the Korean population [39]. This study focused on individuals aged鈥夆墺鈥�65 years in South Korea. Among the 16,119 participants, those aged鈥�<鈥�65 years (n鈥�=鈥�12,795), those with implausible energy intakes (<鈥�500 or >鈥�5,000听kcal) (n鈥�=鈥�335), and those with missing AMD data (n鈥�=鈥�903) were excluded. Finally, a total of 2,086 participants (895 men and 1,191 women) aged鈥夆墺鈥�65 years were included in the analysis. The KNHANES received approval from the Institutional Review Board (IRB) of the Korea Disease Control and Prevention Agency (IRB no. 2018-01-03-P-A). All individuals provided informed consent before enrollment (Fig.听1).

Fig. 1

Flow chart of study participants

General characteristics, anthropometric measurements, and socioeconomic characteristics

Trained staff members conducted participant interviews to collect comprehensive demographic and socioeconomic data. The obtained data included age, BMI, family monthly income, marital status, current smoking status, alcohol consumption, physical activity level, chronic disease status, energy intake, and for women, age at menarche and menopausal status.

Family monthly income was categorized into four groups (low, middle-low, middle-high, and high), while marital status was classified into two groups: 鈥渕arried鈥� and 鈥渓iving as married.鈥� Current smokers were defined as individuals who were presently smokers and had smoked more than five packs of cigarettes in their lifetime. Alcohol consumption was characterized by the consumption of alcohol more than once a month. Physical activity was classified into two categories: 鈥測es鈥� and 鈥渘o,鈥� depending on whether participants engaged in 鈮モ��150听min of moderate physical activity per week, 鈮� 75听min of vigorous physical activity per week, or 鈮モ��150听min of combined moderate and vigorous physical activity per week (with 1听min of vigorous physical activity being equivalent to 2听min of moderate physical activity). Morbidity status was classified as 鈥測es鈥� or 鈥渘o,鈥� based on the diagnosis of one or more chronic diseases, including hypertension, hyperlipidemia, stroke, myocardial infarction or angina, arthritis, asthma, diabetes mellitus, cancer, depression, and renal failure. For women, additional information on menarche age and menopausal status was recorded. BMI was calculated as follows: body weight divided by the square of the height (kg/m虏), providing a measure of body composition.

Assessment of dietary data

Daily food and nutrient intake

Daily food and nutrient intake was evaluated using single 24-h recall data from the KNHANES. Participants gave a detailed reported of the food and beverages they had consumed the previous day via face-to-face interviews. The food items in the dataset were classified into 18 food groups: cereal and cereal products, potatoes and starch, sugar and sugary products, beans and bean products, nuts and seeds, vegetables, mushrooms, fruits, meat and meat products, eggs and egg products, fish and shellfish, seaweed, milk and dairy products, oil and fat, beverages and alcohol, seasoning, processed food, and miscellaneous items. The assessment estimated the intake levels of energy and 16 nutrients, namely, carbohydrate, protein, fat, fiber, sugar, calcium, phosphorus, iron, sodium, potassium, vitamin A, carotene, thiamin, riboflavin, niacin, and vitamin C [30, 40].

aMED scores

Diet quality was assessed using the aMED score. The original Mediterranean diet score, developed by Trichopoulou et al. [41], takes into account the scientific literature on diet and chronic disease risk. In our study, we utilized the aMED score, which was modified by Fung et al. [37]. The aMED score comprises nine components, with a possible total score of 9 points. The intake levels of nine food groups were dichotomized using sex-specific median values as cutoff points. A score of 1 was assigned for consumption above the median level of presumed beneficial foods, such as whole grains, vegetables (excluding potatoes), fruit (including juice), nuts, legumes, and fish, as well as the monounsaturated fatty acid (MUFA)-to-saturated fatty acid (SFA) ratio. Conversely, a score of 1 was also assigned for consumption below the median level of presumed detrimental foods, specifically red and processed meat. A score of 0 was assigned for all other instances. Regarding alcohol intake, 1 point was assigned to men consuming 10鈥�25听g/day and women consuming 5鈥�15听g/day, whereas a score of 0 was assigned for other levels of alcohol consumption.

Assessment of AMD

A comprehensive ophthalmologic examination was conducted by ophthalmologists dispatched by the Korean Ophthalmological Society (KOS) using a mobile unit equipped with ophthalmic devices. The KOS National Epidemiologic Survey Committee provides regular training to these ophthalmologists. Non-mydriatic digital fundus photographs were captured using a TRC-NW6S digital fundus camera (Topcon, Tokyo, Japan), producing a 45掳 retinal fundus image centered on the macula and fovea of each participant鈥檚 eyes. Preliminary grading was performed by trained ophthalmologists using the International Age-Related Maculopathy Epidemiological Study Group grading system. Subsequently, nine retinal specialists, authorized by the KOS, conducted detailed grading. The final grades were determined based on the detailed assessment, and any discrepancies between preliminary and detailed grades were resolved by one retinal specialist. AMD was defined as the presence of soft indistinct or reticular drusen and hard or soft distinct drusen with pigmentary abnormalities.

Statistical analysis

In this KNHANES data-based study, sample weights were generally adjusted for unit nonresponse, and imputation was performed for item nonresponses [42]. Participant characteristics were summarized as the mean and standard deviation (SD) for continuous variables and numbers and percentages for categorical variables using the SURVEY MEAN and SURVEY FREQ procedures, respectively. Participants鈥� aMED scores were divided into three groups: T1 (0鈥夆墹鈥塧MED鈥夆墹鈥�3), T2 (aMED鈥�=鈥�4), and T3 (5鈥夆墹鈥塧MED鈥夆墹鈥�9 for men and 5鈥夆墹鈥塧MED鈥夆墹鈥�8 for women).

One-way analysis of variance was used to determine significant differences in participants鈥� general characteristics according to aMED-score tertile. For the trend test in general linear models, the aMED-score tertile groups were treated as continuous variables based on the median value within each tertile. The covariates included in the adjusted model were selected based on prior literature that highlighted their potential associations with both dietary intake and AMD. Each covariate was either statistically significant in univariate analysis or established in the literature as a risk factor for AMD or diet-related outcomes. The adjusted model accounted for age [9], BMI [10], family monthly income [43], marital status [44], current smoking [18, 45], alcohol consumption [19], physical activity [20, 46], and chronic disease status [14,15,16,17, 47,48,49]. Additionally, in women, age at menarche and menopausal status [50] were specifically adjusted. Multiple regression analysis was performed to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) for AMD according to aMED score via SURVEY LOGISTIC analysis. All statistical analyses were performed using SAS (version 9.4; SAS Institute, Cary, NC, USA). Statistical significance was set at P鈥�&濒迟;鈥�0.05.

Results

Participants鈥� general characteristics according to aMED score

Participant characteristics by aMED-score tertile are displayed in Table听1. In men, the aMED score tended to be higher in participants who were younger (P鈥�=鈥�0.026), had a higher family monthly income (P鈥�=鈥�0.003), consumed alcohol more than once a month (P鈥�=鈥�0.003), were current nonsmokers (P鈥�<鈥�0.001), and engaged in physical activity (P鈥�&濒迟;鈥�0.001). Meanwhile, among women, those with a higher aMED score were younger (P鈥�<鈥�0.001), had a higher family monthly income (P鈥�=鈥�0.026), and experienced menarche earlier (P鈥�&濒迟;鈥�0.001).

Table 1 Participants鈥� general characteristics according to aMED-score tertile

Daily food and nutrient intake according to aMED score

The daily food intake of Korean older adults by aMED-score tertile is presented in Tables听2 and 3. In men, daily food intake, excluding that of sugar and sugary products, mushroom, meat and meat products, milk and dairy products, beverages and alcohol, processed food, and others, generally exhibited a significant increase among those with higher aMED scores (P-迟谤别苍诲鈥�&濒迟;鈥�0.05). Similarly, daily food intake among women, except that of potatoes and starchy products, sugar and sugary products, mushroom, meat and meat products, milk and dairy products, beverages and alcohol, processed food, and others, was significantly higher in those with higher aMED scores (P-迟谤别苍诲鈥�&濒迟;鈥�0.05). Generally, participants with higher aMED scores displayed significantly increased daily food intake, except for that of alcohol (P-迟谤别苍诲鈥�&濒迟;鈥�0.05).

Table 2 Participants鈥� daily food intake according to aMED score

Table 3 Participants鈥� daily nutrient intake according to aMED score

Association between aMED score and AMD

The risk of AMD by aMED-score tertile is presented in Table听4. Men with the highest aMED scores (5鈥�9 points) exhibited a significantly reduced AMD risk compared with those in the lower aMED-score tertiles after adjusting for various factors, such as age, BMI, family monthly income, current smoking, alcohol consumption, physical activity, chronic disease status, and energy intake (adjusted OR鈥�=鈥�0.58; 95% CI鈥�=鈥�0.39鈥�0.88; P-trend鈥�=鈥�0.021). In women, confounding factors such as age at menarche, menopausal status, alcohol consumption, smoking, income, marital status, diagnosis of disease, and engagement in aerobic exercise were considered. However, no significant association was observed between the aMED score and AMD in women (adjusted OR鈥�=鈥�0.88; 95% CI鈥�=鈥�0.61鈥�1.29; P-迟谤别苍诲鈥�=鈥�0.691).

Table 4 Odds ratios with 95% confidence intervals of aMED-score tertiles and AMD

Discussion

This study investigated the relationship between overall diet quality, as indicated by the aMED score, and the prevalence of AMD using 2017鈥�2018 KNHANES data. Our findings revealed a statistically significant association between a higher aMED score and lower AMD prevalence among Korean older men (adjusted OR鈥�=鈥�0.58; 95% confidence interval鈥�=鈥�0.39鈥�0.88; p-trend鈥�=鈥�0.021). This cross-sectional study鈥檚 results, in which older men in the third aMED score tertile had a 42% lower AMD prevalence than those in the first tertile, are consistent with those of a cohort study that found a greater adherence to the Mediterranean diet to be associated with a 41% reduced risk of incident advanced AMD [51]. To the best of our knowledge, this study is the first to demonstrate a positive association between the aMED score and a low risk of AMD in the Korean older population.

The aMED scoring system provides a comprehensive evaluation of dietary habits, considering both beneficial and detrimental components of the diet. Originally developed and extensively validated in Western populations, the Mediterranean diet emphasizes a high consumption of plant-based foods, such as fruits, vegetables, legumes, nuts, whole grains, and olive oil, while encouraging a moderate intake of fish, poultry, and dairy and low consumption of red meat [31]. Given the differences in dietary patterns between Western and Korean populations, this study applied a modified aMED score [37] to better reflect Korean dietary habits while maintaining the core principles of the original aMED score. For instance, since olive oil is not a staple in traditional Korean cuisine, the inclusion of healthy fats was adjusted to consider the fat types more commonly consumed in Korea, such as fish oil and plant-based oils. Similarly, while whole grains and legumes are prevalent in both diets, specific food types, such as soy-based products (e.g., tofu, soybean paste, etc.), were integrated into the adapted score to align with Korean dietary customs. This score ensures that the assessed dietary quality remains valid and relevant to this population by incorporating the Mediterranean diet鈥檚 main components while adapting it to Korean staples. This approach presents a holistic view of diet quality while respecting regional variations. The components of the aMED score, which include a high consumption of fish, legumes, fruits, and vegetables, effectively align with dietary patterns previously associated with a lower AMD risk. Several studies have reported an association between dietary components similar to those included in the aMED and a reduced AMD prevalence. Kim et al. [27] found increased fruit and vegetable consumption to provide some protection against AMD. Kang et al. [30] demonstrated that a dietary pattern rich in fish and legumes, among the 11 food groups, has a protective effect against AMD in the Korean population. In the cross-sectional, population-based European Eye Study, Augood et al. [52] reported that oily fish intake (鈮モ��1 time/week vs. < 1 time/week) was associated with the risk of neovascular AMD. Additionally, Chiu et al. [53] found that a compound score consisting of vitamins C and E, zinc, lutein/zeaxanthin, omega-3 fatty acids, and a low dietary glycemic index positively influenced the risk of early AMD. Their analysis focused on individual food components, while our study used the aMED score, which evaluates overall diet quality by consolidating multiple food group scores into a single comprehensive score. Merle et al. [51] revealed a positive association between high Mediterranean diet scores and a reduced risk of AMD progression, exhibiting consistency with our study, by using dietary quality scores, such as the aMED score, instead of analyzing isolated dietary components, thereby providing a more holistic perspective on the association between diet and AMD.

Several mechanisms potentially underlie the association between AMD and the aMED score. In our study, the aMED score was calculated based on components such as whole grains, nuts, legumes, fruits, vegetables, red and processed meat, fish, alcohol, and the MUFA: SFA ratio [37]. Previous research indicates that antioxidants play a crucial role in macular degeneration, with high Mediterranean diet scores being associated with an elevated glutathione: glutathione disulfide ratio [32]. Consequently, high-dose antioxidants have been linked to a decelerated progression of AMD [54, 55]. Therefore, our study suggests that adherence to a Mediterranean diet potentially exerts beneficial effects on AMD by reducing oxidative stress. Moreover, polyunsaturated fatty acids (PUFAs) play a protective role against AMD [56]. Dietary PUFA intake is associated with anti-inflammatory properties, generating anti-inflammatory mediators, such as resolvins and protectins. A high omega-3 fatty acid intake has been linked to reduced levels of serum C-reactive protein and other pro-inflammatory mediators [56,57,58]. Another potential mechanism involves the effects of a high dietary salt intake and low water consumption. Increased plasma osmolality potentially exacerbates age-related retinal diseases by stimulating local inflammation and angiogenic factors, such as vascular endothelial growth factor, placental growth factor, and basic fibroblast growth factor, as well as the expression of aquaporin-5 production in the retinal pigment epithelium [59].

In our study, the aMED score was associated with a lower OR of AMD in men but not in women. The reasons for this sex-based difference remain unclear. Similar studies by Rudnicka et al. [11] and the AREDS Research Group [14] have suggested several possible explanations for these sex-based differences. According to these studies, women may possess lower statistical power to detect a difference or may have a history of the disease. Sex hormones [60, 61] may also influence the prevalence of AMD in women. In addition, differences in food intake between men and women are noteworthy, as the aMED score is based on overall dietary patterns rather than specific food groups or nutrients. Further research is required to elucidate the mechanisms underlying these apparent sex-based differences in the association between the aMED score and AMD.

This study has several limitations. First, as the KNHANES is a cross-sectional study, establishing a causal relationship between the aMED score and AMD proved impractical. Second, our reliance on a single 24-h dietary recall might not have entirely captured participants鈥� usual dietary intake. Third, we did not account for dietary AREDS-related supplement intake. Therefore, a limitation exists in not providing additional insight into the potential impact of supplement-based antioxidant nutrient intake on AMD risk, which is an area warranting improvement in future studies. Fourth, this study relied on non-mydriatic digital fundus photography for AMD diagnosis, despite some participants in the older population potentially having cataracts or small pupils, which could have rendered it difficult to capture clear retinal images and might have affected the accuracy of AMD grading. Additionally, certain participants were excluded owing to refusal or inability to obtain retinal images, potentially introducing selection bias. Additionally, we were unable to assess AMD by stage (early, intermediate, or advanced) or type (wet or dry) owing to data limitations. Moreover visual acuity, which could have provided further insight into the functional impact of AMD, was not measured. Finally, selection bias might have emanated from the exclusion of participants with missing KNHANES data, potentially resulting in overestimation or underestimation.

Despite these limitations, this study has several strengths. It utilized a nationally representative sample of older Koreans and is the first to investigate the association between the aMED score and AMD risk in this population. The use of the aMED score, which evaluates overall diet quality, enabled a holistic assessment of dietary patterns rather than focusing on individual nutrients or food groups.

Conclusions

In conclusion, our findings suggest that higher adherence to a Mediterranean-style diet, as reflected by a higher aMED score, is associated with a lower prevalence of AMD among older Korean men. However, considering the cross-sectional nature of this study, further research using longitudinal or interventional designs is required to better understand the potential protective role of diet quality in AMD prevention.