Another paper with HapMap data, Allele Frequency Matching Between SNPs Reveals an Excess of Linkage Disequilibrium in Genic Regions of the Human Genome in PLOS Genetics (so totally open access). The synopsis is clear, so no need to quote. Two points

1) This is a methodological attempt to attain more precision in ascertaining linkage disequilibrium within the genome in regards to SNPs by comparing those variants where frequencies matched.

2) The empirical results are the same as previous papers. a) Evidence of massive selection via "haplotype blocks," basically signatures of selective sweeps through human populations, and, b) more of these in Europeans and East Asians than Africans.

If you get confused when we talk about LD, read the paper. It isn't all that exciting, but I think the repetition and banality of the exposition would be beneficial if you're unclear in regards to the concept.

Update by Darth Quixote: The PLoS site has an Excel file listing the genes contributing most to the excess LD in the respective samples. Perhaps our genomicist friends will get something of it.