The ongoing 12-year Connecticut Longitudinal Study, involving a representative sample of 445 schoolchildren, has found that in typical readers, IQ and reading not only track together, but also influence each other over time. But in children with dyslexia, IQ and reading are not linked over time and do not influence one another. Although this difference has been assumed, this is the first direct evidence for it. It should also be noted that the language problem is not confined to reading: those with dyslexia take a long time to retrieve words, so they might not speak or read as fluidly as others.
(2010). Uncoupling of reading and IQ over time: empirical evidence for a definition of dyslexia.
Psychological Science: A Journal of the American Psychological Society / APS. 21(1), 93 - 101.