Implementing a 'read-through' caching strategy will minimize read latency and reduce the load on the DynamoDB table by caching the data after the first read from the database. Subsequent reads are served directly from the cache, which decreases the burden on the table. As the cache is populated on-demand during a database read, it ensures that the data in the cache is consistent with the database. Other caching strategies might not be as effective for this scenario. For example, write-through caching is more suited for write-heavy applications, lazy loading can lead to stale data if not managed properly, and TTL doesn't directly address the problem of high read load and throttling.