We tested 503,984 SNPs for association with hypertension status and BP traits. Our study is the first report of a combined genome-wide linkage and association scan for these traits in families with Chinese ancestry. The two traits are correlated and heritable, and SBP shows stronger increases with age, with DBP starting to plateau and in some individuals fall at ages above 60–65 years 
. Some 
have suggested the study of mean arterial pressure, which increases with advancing age and is highly correlated with SBP and DBP, also showed evidence of heritability. In our GWAS, we chose to examine SBP, DBP and MAP as separate traits, and also carried out an association analysis of hypertension disease status.
We report significant genome-wide linkage as well as an association of common genetic variants with hypertension. Our linkage analysis revealed two novel regions of linkage for DBP, 5p13.1 and 5p13.2, formally achieving genome-wide significance, and also one novel region for SBP, 2q22.3 (LOD
3.01). Our association analysis has identified a significant variant (rs6596140) for hypertension and also detected effects of relatively common alleles with modest P
values for quantitative traits. The region of association () is limited to a 3-kb region flanked by strong recombination hotspots, which are surrounded by multiple non-coding sequences but no known genes. FSTL4
, which is a key modulator in muscle development, was specified as nominally associated with increased risk of stroke in a cardiovascular health study 
. It is possible that the nearby associated SNP we detected plays a regulatory role for FSTL4
There is limited overlap between the regions of strongest linkage and association. However, further assessment of SNP association under the linkage peak on chromosome 5 did yield a significant variant, rs1605685, after correcting for multiple testing of the 808 SNPs under the peak.
Both linkage and association analysis provide useful, but different forms of information in identifying genetic contributions to complex traits. Linkage analyses are used with family data to find broad genomic regions that contain putative disease loci, while association analyses can identify much smaller regions, either a causal variant or one which is in linkage disequilibrium with such a disease-causing locus. While association can generally detect smaller effects than linkage, association is limited to detecting variants that are either directly assayed or are in strong LD with typed SNP. These are more likely to be common variants, which generally have small effects. In contrast, linkage in families can detect rare variants specific to a subset of families and these rare variants are likely to have larger effects. An advantage of family data in association analyses is to provide a perfect control for population stratification. Our combined linkage and association design brings the best of both worlds to the problem at hand. If significant linkage is detected in the presence of significant association, it suggests that the putative locus is not the functional gene, but rather is a locus in disequilibrium with a trait locus. In contrast, if the linkage signal disappears at a point of significant association, this is suggestive that the associated SNP is, in fact, causal. In addition, all our subjects are Hong Kong Han Chinese, a relatively homogeneous group with regard to genetic background and environmental risk factors. Nevertheless, our findings are limited by relatively small sample size. Furthermore, as prior studies have reported ethnic differences in frequencies of alleles and effects of genes involved in blood pressure traits 
, the novel loci we found may not be generalized to other population groups and await further replication.