Details of AMD characteristics are shown in Table 1. Of the 3,127 individuals, 2,825 (90.34%) were classified as having no AMD, while 302 (9.66%) showed any form of AMD. Specifically, 274 individuals (8.76%) had early AMD, while 28 (0.90%) were diagnosed with late AMD. The total prevalence of AMD in any form was 9.66%, with an average participant age of 71.248. Non-Hispanic White was the most common ethnicity (71.192%). Compared to the no AMD group, participants with AMD had higher indices of WT, CI, ABSI, and WWI (P < 0.05). However, gender, race, and education were not significantly different between the any AMD and without AMD groups (P > 0.05). Unmarried individuals exhibited a markedly elevated prevalence of AMD, particularly in the late stage (P < 0.05).

Regarding lifestyle factors, no substantial differences in drinking and smoking were identified across any of the stages of AMD (P > 0.05). Regarding health conditions, individuals with AMD were notably more likely to report a history of CVD and hypertension (P < 0.05). Additionally, more individuals with AMD had a history of cataract surgery or glaucoma (P < 0.05). There were no substantial variations in the prevalence of diabetes among the various stages of AMD (P > 0.05).

The characteristics of the participants are expressed as means with 95% CI for continuous variables and as percentages with 95% CI for categorical variables. P-value presented is based on the comparison between participants without AMD and those with any AMD.

Table 2 provides a summary of the weighted logistic regression analysis results for any form of AMD. In unadjusted model 1, a higher ABSI Z-score (OR = 1.482 (1.306−1.681), P < 0.0001), CI Z-score (OR = 1.345 (1.182−1.531), P = 0.0001), and WWI Z-score (OR = 1.492(1.299−1.714), P < 0.0001) were associated with higher odds of any AMD. After full adjustment, the WtHR Z-score was markedly correlation with any AMD (OR = 1.237 (1.065−1.438), P = 0.0215). Additionally, the BRI Z-score revealed a significant association with any AMD (OR = 1.221(1.058−1.410), P = 0.0232). The CI Z-score maintained its significance (OR = 1.189 (1.039−1.362), P = 0.0334), and the WWI Z-score retained its association with any AMD (OR = 1.250 (1.095−1.425), P = 0.0090).

Subsequently, we investigated whether these anthropometric indices impacted the severity of AMD. The results of the association between anthropometric indices and early AMD are provided in STable 1. After full adjustment, the WtHR, BRI, CI, and WWI exhibited significant associations with early AMD (OR = 1.269 (1.089−1.479), P = 0.0138; OR = 1.251 (1.081−1.449), P = 0.0151; OR = 1.209 (1.045−1.399), P = 0.0310; and OR = 1.281 (1.116−1.471), P = 0.0064). As for late AMD (STable 2), all indices did not exhibit significant associations across models. These results suggest that these anthropometric indices (WtHR, BRI, CI, and WWI) can be important indicators for early AMD; however, their relevance is reduced for late AMD.

Weighted logistic regression analysis showing the correlation between anthropometric indices and any AMD. Model 1: unadjusted model. Model 2: adjusted for gender, age and race. Model 3: adjusted for gender, age, race, marital status, education, drinking, smoking, hypertension, diabetes, CVD, cataract operation and glaucoma.

Fig. 2-5 shows the association between WtHR, CI, BRI, WWI, and AMD by stratified multivariate logistic regression analysis across various demographic and health-related subgroups, including gender, age, BMI, drinking, smoking, CVD, diabetes, hypertension, cataract operation, and glaucoma. There was a significant positive connection between WtHR and AMD in individuals who were overdrinking (P for interaction = 0.0033) and did not smoke (P for interaction = 0.0402). However, the correlation between WtHR and AMD did not vary significantly across age, gender, BMI, CVD, diabetes, hypertension, cataract surgery, and glaucoma subgroups (Fig. 2). Furthermore, CI, BRI, and WWI exhibited a significant positive correlation with AMD in participants with overdrinking (P for interaction = 0.0194, 0.0021, and 0.0022). The WWI and CI were significantly associated with AMD with females (P for the interaction = 0.0146 and 0.0117, respectively). In contrast, the BRI demonstrated a consistent relationship with AMD across both genders (Figs. 3−5). The relationships between CI, BRI, and WWI with AMD remained consistent across multiple subgroups, including divisions by age, BMI, smoking habits, drinking habits, presence of CVD, diabetes, hypertension, history of cataract surgery, and glaucoma.

Subgroup analyses to determine the correlation of WtHR and AMD. Multivariable weight logistic analyses were conducted after adjusting for age, gender, BMI, drink, smoke, CVD, diabetes, hypertension, cataract operation, glaucoma

Subgroup analyses to determine the correlation of CI and AMD. Multivariable weight logistic analyses were conducted after adjusting for age, gender, BMI, drink, smoke, CVD, diabetes, hypertension, cataract operation, glaucoma

Subgroup analyses to determine the correlation of BRI and AMD. Multivariable weight logistic analyses were conducted after adjusting for age, gender, BMI, drink, smoke, CVD, diabetes, hypertension, cataract operation, glaucoma

Subgroup analyses to determine the correlation of WWI and AMD. Multivariable weight logistic analyses were conducted after adjusting for age, gender, BMI, drink, smoke, CVD, diabetes, hypertension, cataract operation, glaucoma

To illuminate the correlation between WtHR, CI, BRI, and WWI with AMD, smooth curve fitting model was employed. A two-segment linear regression model analysis revealed nonlinear associations for WtHR, BRI, and WWI with AMD (inflection points at 0.535, 4.045, and 12.065, respectively). CI exhibited a linear and positive correlation with AMD (Fig. 6).

Smooth curve fitting models evaluated the correlation between four anthropometric indices with AMD. Adjusted smooth curve fitting models adjusted for gender, age, race, marital status, education, drinking, smoking, hypertension, diabetes, CVD, cataract operation, and glaucoma. The red line illustrates the smoothed curve that fits the data points, while the blue shaded areas indicate the 95% CI around the fit. (A) Smooth curve fitting model of WtHR. (B) Smooth curve fitting model of CI. (C) Smooth curve fitting model of BRI. (D) Smooth curve fitting model of WWI