Worldwide, nearly 2.2 billion individuals face vision impairment or blindness, mainly caused by age-associated eye disorders like cataracts, diabetic retinopathy, and glaucoma. If these ailments are not recognized and treated timely, they may lead to irreversible blindness. The research aims to develop a highly accurate and efficient model for the early prediction of these eye conditions by using fundus pictures. Existing models for detecting eye diseases have made significant progress, but many still fail to provide the necessary accuracy for early detection, which is crucial for efficient treatment. EDCNet seeks to address this challenge by integrating CNNs and GRUs in a hybrid model. CNNs are employed for their strength in feature extraction, capturing intricate details from fundus images. At the same time, GRUs are used for their ability to handle sequential data, enabling the model to analyze temporal patterns in the retinal images. The model’s preprocessing phase is enhanced to increase the quality and clarity of the dataset, which is vital for accurate analysis. EDCNet utilizes the Adam optimizer, known for its efficiency in training deep learning models, to optimize the model’s performance further. This combination of advanced techniques aims to create a reliable system for detecting early signs of age-related eye diseases, facilitating timely intervention and potentially reducing the global burden of blindness. The proposed work aims to contribute to more accurate and accessible eye disease screening, especially in regions where early detection can significantly impact public health outcomes. The EDCnet model has been evaluated and exhibits a high level of accuracy, achieving a precise performance rate of 91.75%.