MATERIALS AND METHODS We have recruited 59 nuclear families (a total of 214 subjects) from a hypertension clinic at Taipei Veterans General Hospital, Taiwan. Our study group included 81 young onset hypertensive patients (59 probands and 22 affected sibs, mean age 30.4 (SD 0.95)), 39 normotensive sibs (mean age 32.2 (SD 1.6)), and 94 parents. Our previous study included 25 affected sib pairs from 18 families for affected sib pair analysis. This transmission disequilibrium test (TDT) study used information from all 59 families with probands. Therefore, the former is a subset of the latter. The protocol of this study was approved by the Human Investigation Committee of the Institute of Biomedical Sciences, Academia Sinica. Polymorphic microsatellite markers located on 1q42-43 and 17q23 were selected based on GeneMap’99 and comprehensive human genetic maps from the Marshfield Medical Research Foundation, and obtained from Multi-Colored Fluorescent Human MapPairs Markers of Research Genetics (Huntsville, AL). Nine markers on 1q42-43 were selected: D1S2805 (245.05 cM), D1S3462 (247.23 cM), D1S459 (247.23 cM), D1S1540 (252.12 cM), D1S235 (254.64 cM), D1S517 (262.96 cM), D1S1149 (262.96 cM), D1S1594 (265.49 cM), and D1S547 (267.51 cM). The six markers on 17q23 were D17S1297 (83.40 cM), D17S1295 (83.40 cM), D17S942 (85.94 cM), ATA108a05 (88.76 cM), D17S789 (89.32 cM), and D17S2193 (89.32 cM). The polymerase chain reaction protocol for microsatellite markers was performed as previously reported. 3 Fragment analysis was performed using an ABI 377 DNA sequencer and analysed by GeneScan version 3.0 and GenotypeR version 3.0. The allele calling was conducted independently by two readers and cross checked. In addition to conventional TDT that only used data from heterozygous parents, an extended TDT (S-TDT and combined Z score) developed by Spielman et al4 was also carried out. Information from 42 young hypertensive patients and both of their parents (representing 35 families) were used for conventional TDT. In the sibship transmission disequilibrium test (S-TDT), information from 25 hypertensive patients and 21 normotensive sibs from 17 families were used. Then a Z score was obtained by combining information from the conventional TDT and the S-TDT. For comparisons, we also carried out Horvath’s SDT, 5 which uses an exact p value to test the difference between affected and unaffected sibs. In this test, only information from 44 young hypertensive patients and their normotensive sibs (39) were used. Since these markers were close to each other, we performed haplotype TDT using the “TRANSMIT” program. 6 In this analysis, data from all 59 families were used. Information from the 24 single parent families was incorporated into the analysis by use of expectation maximisation algorithms. Because of our limited sample size, only haplotypes created by two markers were included. Because we tested 15 markers with multiple alleles, a Bonferroni procedure was carried out to adjust for multiple comparisons, as suggested by Spielman et al. 4

RESULTS AND DISCUSSION No association with any marker in the region of 1q42-43 was found (data not shown). The AGT was the first gene linked to hypertension, 7 but results showing linkage of AGT and hypertension has not been consistent across various populations. A