Hi Sailesh, intronic SNPs can potentially influence splicing. As I'm sure you can imagine, this can take a few different forms (large insertion / frameshift) and the impact on the protein (if it's still synthesised) can be dramatic. Therefore, an intronic SNP can reduce or remove protein function and, depending on the gene in question, has the potential to be causative in disease. I think there are also some cases where introns towards the 5' end of a gene contain some regulatory elements whose mutation might impact on the rate of transcription... but you might need to look that up yourself... or perhaps someone else will enlighten us both!
In addition, if the mRNA from the given gene may undergo alternative splicing, this intron can be included into alternative form of this protein. Then this SNP may affect the structure and function of this protein. Moreover, introns may have their own transcription units producing regulatory RNAs or perhaps small proteins in that cases SNP may affect function of the corresponding products of intron expression.
For an association study you are looking for haplotypes that might have higher or lower risc for a certain disease. So it is not important whether these SNP's itself alters the function of that gene in which the SNP is located, but it can be linked to another SNP in this haplotype that does change the function of another gene or a set of genes in the naberhood.