Simon Garfield has an interesting piece at the Observer Magazine on dyslexia and its treatment (here's a longish chunk from the middle):
The classic symptoms - difficulty in reading, writing or spelling among those who otherwise possess an average or high level of intelligence - naturally led educators to believe it was solely a linguistic problem, and there was no reason to search for automatic correlation with impairments in the brain. The phonological theory, which states that reading problems are due to children not detecting the correct sound of written letters and words, is still pre-eminent, but the refinement of brain-scanning techniques and genetics has established beyond doubt that there are often significant differences in the brain between those who are dyslexic and those who are not. The core of John Stein's research has been devoted to showing what causes these differences, and in so doing suggest potential advances in early diagnosis and treatment.
Some of this research is at a primitive level due to limited funds, and cohort studies are small. Stein's clinic has had good results with the use of coloured lenses in reading glasses (about a third of the 500 children assessed by the DRT each year show improvements with blue or yellow lenses), and an increased intake of fish oils rich in omega-3 benefits another third (possibly because omega-3s can improve the function of the magnocellular systems in the brain that help to stabilise visual perception).
Some of the research being done at Oxford is backed up by large international studies in Europe and the United States, particularly the genetic work. It is now accepted that over half of the differences in children's reading is due to genetic factors inherited from their parents. Inevitably, the hunt is on for the specific gene that may identify this predisposition, and Stein's colleagues have identified a gene on chromosome 6, known as KIAA0319, a key factor in the way the brain develops. When this gene is removed in mice, cell growth in the brain is reduced; a similar, though milder, deficit is visible in dyslexic brains post-mortem.