“…The complexity and risk of confounding under such con-ditions has encouraged prospective studies of firstepisode patients in early or prodromal phases of illness. Most first-episode studies involve patients diagnosed with early (Jones & Tarrant, 1999;McGorry et al, 2000;Cannon et al, 2001;Gaebel et al, 2001;Hollis 2003), or established schizophrenia (Kane et al, 1982;Biehl et al, 1986;Schubart et al, 1986;McCreadie et al, 1989;Johnstone et al, 1990;Tohen et al, 1990b;1992a;1996;Tohen 1991;Leff et al, 1992;Ram et al, 1992;Ventura et al, 1992;Lieberman et ai, 1993;Bromet et al, 1996;Varma et al, 1996;Craig et al, 1997;Gupta et al, 1997;Lay et al, 1997). There have been far fewer first-episode follow-up studies of patients diagnosed with bipolar disorder (Tohen et al, 1990b(Tohen et al, , 2000aFennig et al, 1996;Strakowski et al, 1998;Conus et al, 2004;Schimmelmann et al, 2005), or other types of psychotic disorders (Pillmann et ai, 2002;Schimmelmann et ai, 2005;Abe et ai, 2006;Emsley et ai, 2006;2007).…”