On the basis of the associations between mouse H-2 types and immune response, many studies were carried out on the associations between HLA types and diseases with a possible immune etiology. Early data are summarized in Figure 1. These studies were simple case–control comparisons of the frequencies of different HLA types in disease as compared to control populations. The most notable early result was the association between HLA-B27 and ankylosing spondylitis. Of the diseases shown in Figure 1, the only one with no connection with an immune etiology is hemochromatosis, which became the first and possibly still the best example of finding, by LD, a previously unknown functional gene for a relatively common disease12. The ORs for most of the ten or more diseases that had been investigated in several different studies by 1974 were above 5, with that for ankylosing spondylitis being over 100. The corresponding χ2 values were nearly all at least 15, and many were much greater. The exceptions to high ORs were those for multiple sclerosis (OR = 1.7), acute lymphatic leukemia (OR = 1.7) and
