There is probably no other non-lethal health problem except perhaps allergies that afflicts so many breeds of dogs as hip dysplasia. It cripples dogs with pain, sometimes in the prime of their lives, and there is very little modern veterinary care can do about it. It seems clear that it has some genetic component (it is thought to be polygenic) but there are clearly environmental (i.e., non-genetic) influences as well. There has been some modest success in reducing its incidence in some breeds by screening programs, but for the most part it remains an intractable problem and the focus of many research programs.
Consequently, I was quite surprised to run across a paper (1) published in 2006 about a study that was able to substantially reduce in incidence and severity of hip dysplasia in Labradors – not by locating particular genes or implementing strategically-designed breeding programs – but by reducing food consumption. (Download a copy here - pdf) Conducted by Nestle Purina in collaboration with a slate of veterinarians and academics, the study used 48 Labrador Retriever puppies from 7 litters. In each litter, puppies were paired and one assigned to the control group and one to the treatment group. The control group was provided food ad libitum (unrestricted) starting at 8 weeks, and each puppy in the treatment group was fed 25% less than the amount consumed by its pair in the control group. Their weight was monitored and hips x-rayed at regular intervals throughout the lifetimes of the dogs.