Meta-analysis Report

Reference	Lasky-Su, J. A.,2005 PMID: 15578606
Citation	Lasky-Su, J. A., S. V. Faraone, et al. (2005). Meta-analysis of the association between two polymorphisms in the serotonin transporter gene and affective disorders. Am J Med Genet B Neuropsychiatr Genet 133B(1): 110-115.
Disease Type	Bipolar Disorder & Major Depressive Disorder
Study Type	Candidate-gene association study

Samples	For 44-bp Polymorphism,BP:a total of 2,774 cases and 3,652 controls in 15 studies,UP:1,961 cases and 3,402 controls in 14 studies;for 17-bp VNTR,BP:2,292 controls and 1,357 cases in 11 studies,UP:1,817 controls and 653 UP individuals in 9 studies
Statistic Method	44-bp Polymorphism:We summarized the strength of association in these two-by-two tables using the odds ratio (OR);17-bp VNTR polymorphism:Likelihood ratio tests (LRTs) comparing two logistic regression models were used to determine if the alleles were best analyzed categorically or continuously.Each group of studies was analyzed by random-effects metaanalysis.ORs were pooled according to the methods of DerSimonian and Laird [1986].The significance of the pooled OR was determined through the z-test. The heterogeneity of the group of ORs was assessed using a X2 test of goodness of fit.The influence of individual studies on the pooled OR was determined by sequentially removing each study and recalculating the pooled OR and 95% CI. We also assessed publication bias using the method of Egger et al.[1997].In all analyzes, Bonferroni corrections were made by setting the type-I error rate, alpha, to 0.025.
Basic Result	1.44-bp Polymorphism:The short allele(s) of the 44-bp insertion/deletion polymorphism showed a significant association for BP (odds ratio but not UP. for BP,the combined OR was significantly different than 1.0, with an estimate of 1.13(z=3.22, P=0.001, CI (OR): [1.05-1.22]). In addition, the findings were not influenced by any one study, there was no evidence for heterogeneity of the ORs (X2=13.02, df=14, P=0.53; ethnicity effect P=0.16), and there was no evidence for publication bias (t=0.20, P=0.84).For UP,the combined OR estimate was not significantly different than 1.0 (OR=1.05, z=1.07, P=0.28,CI (OR): [0.96-1.14]). There was significant evidence for heterogeneity of the ORs (X2=26.16, df=13; P=0.02); however,this could not be explained through and ethnicity effect(z= 1.35, P=0.18).2.17-bp VNTR:an increase in the number of tandem repeats had no significant association with any of the disorders. for BP, the combined OR for these 10 studies resulted in an OR that was not significantly different from 1.0 (OR=1.05, z=1.06, 95% CI (OR): [0.96, 1.14], P=0.29).This analysis indicated that there was significant heterogeneity among the studies (X2=22.23, df=10, P=0.01) that could not be explained by ethnicity (z=0.42, P=0.67) or diagnostic instrument (z=1.06, P=0.29) and there was insufficient information to test whether other variables may have contributed to this heterogeneity.There was also no evidence for publication bias (t= 0.37, P=0.72) and no one study impacted the findings substantially.For UP,The combined analysis yielded an overall OR that was not significantly different than 1.0, with an estimate of 0.99(z= 0.29, P=0.77, CI (OR): [0.92-1.06]). A sensitivity analysis,in which the combined estimate of the OR was computed after omitting one study at a time, showed that no one study influenced these findings substantially. There was also no evidence of publication bias (t=0.17, P=0.87) or heterogeneity(P=0.98).