Table 1.
Studies Differently Assessed in Three Different Meta-Analyses
| Study | Jönsson et al., 2003 | Glatt et al., 2003 | Present study | Notes |
|---|---|---|---|---|
| Arinami et al., 1994 | Included* | Included | Included | See footnote. |
| Arinami et al., 1996 | Included* | Included | Included | Among patients, 258 Ser alleles were used in Jönsson et al. (2003) and the present study. Glatt et al. (2003) used 260 Ser alleles based on a typographic error in the report. |
| Fujiwara et al., 1997 | Not included | Included | Included | Study was unavailable to Jönsson et al. (2003). |
| Gejman et al., 1994 | Included | Included | Included | Among controls, 246 Ser alleles were used in Jönsson et al. (2003) and the present study. Glatt et al. (2003) used 222 Ser alleles. |
| Goldman et al., 1997 | Included | Not included | Not included | Subjects had a primary diagnosis of alcohol abuse or dependence, not schizophrenia. |
| Harano, 1997 | Included | Included | Included | Among patients, 132 Ser alleles were used in Glatt et al. (2003) and the present study. Jönsson et al. (2003) used 130 Ser alleles. |
| Itokawa et al., 1993 | Not included | Included | Not included | Complete overlap with Arinami et al. (1994) was verified by the corresponding author. |
| Jönsson et al., 2003 | Included | Not included | Included | Study was unavailable to Glatt et al. (2003). |
| Morimoto et al., 2002 | Included | Not included | Included | Genotype frequencies were unavailable to Glatt et al. (2003). |
| Nanko et al., 1994 | Not included | Included | Not included | Complete overlap with Hattori et al. (1994) was verified by the corresponding author. |
| Ohara et al., 1996 | Not included | Included | Included | Study was unavailable to Jönsson et al. (2003). |
| Sasaki et al., 1996 | Included | Included | Included | Among controls, 499 Ser and 11 Cys alleles were used in Glatt et al. (2003) and the present study. Jönsson et al. (2003) used 498 Ser and 10 Cys alleles. |
| Shaikh et al., 1994 | Included | Included | Included | Among patients, 287 Ser alleles were used in Jönsson et al. (2003) and the present study. Glatt et al. (2003) used 273 Ser alleles. |
| Spurlock et al., 1998: Austria | Included* | Included* | Included | See footnote. |
| Spurlock et al., 1998: Ireland | Included* | Included* | Included | See footnote. |
| Spurlock et al., 1998: Italy | Not included | Included* | Not included | Overlap with Verga et al. (1997) was suspected, but could not be confirmed; thus, to be conservative, the Italian sample was excluded in the present study. |
| Spurlock et al., 1998: Sweden | Included* | Included* | Included | This sample was pooled with the Swedish subjects of Jönsson et al. (2003), while Glatt et al. (2003) pooled this sample with the other samples from Spurlock et al. (1998). The sample was included in the present meta-analysis as a separate study. |
| Spurlock et al., 1998: Wales | Included* | Included* | Included | See footnote. |
*
In Jönsson et al. (2003), subjects from the same research centers were pooled and considered as one study (e.g., the two studies of Arinami et al. (1994, 1996) were pooled), while Glatt et al. (2003) treated these as semi-independent reports with the overlapping samples excluded. In the collaborative study of Spurlock et al. (1998), samples from different centers were separated and each was treated as an independent study in Jönsson et al. (2003), while Glatt et al. (2003) treated all samples in Spurlock et al. (1998) as one study.