Major depression is widely acknowledged as a major health problem, with adverse consequences in terms of loss of productivity and lowered quality of life (Ebmeier, Donaghey, & Steele, 2006). Several psychological treatment options exist showing fairly equivalent outcomes (Cuijpers, van Straten, Andersson, & van Oppen, 2008), and evidence suggests that psychological treatments for mild to moderate depression are about equally effective as pharmacological treatments (Cuijpers, van Straten, van Oppen, & Andersson, 2008). When equal effects are found other aspects become relevant when making decisions about treatments for depression. One issue concerns patient preferences, as patients often prefer psychological treatments (Leykin et al., 2007 and van Schaik et al., 2004), even if treatment preferences per se do not need to have an impact on outcome ( Leykin et al., 2007). Another issue concerns costs and cost-effectiveness of the treatments ( Hargreaves, Shumway, & Hu, 1999). Yet another concern relates to access to the treatment. With increasing demand for psychological treatments, guided self-help approaches have been developed showing promising outcomes ( den Boer, Wiersma, & Van den Bosch, 2004). A recent promising and potentially cost-effective treatment format involves delivering cognitive behaviour therapy (CBT) over the Internet ( Andersson, 2009 and Barak et al., 2009). Internet-delivered CBT for major depression has been tested in a number of trials, with varying outcomes from promising ( Andersson et al., 2005, Christensen et al., 2004, Kessler et al., 2009, Meyer et al., 2009, Perini et al., 2009 and Spek et al., 2007), to somewhat less effective or not effective at all effective against no treatment or treatment as usual control groups ( Clarke et al., 2005, Clarke et al., 2002, de Graaf et al., 2009 and O'Kearney et al., 2006). The effects of Internet-delivered CBT were summarized in a meta-analysis showing a moderate between-group effect size of Cohen's d = .40 ( Spek, Cuijpers et al., 2007). However, the same authors and others noted that self-help programmes in which support was provided (d = .61) were more effective than programmes without support (d = .25) ( Andersson, 2006 and Andersson and Cuijpers, 2009), and more recent trials support this observation ( de Graaf et al., 2009). There may also be other differences between the guided and unguided treatments. For example, the guided treatments tend to be more comprehensive and longer. Indeed, one exception is a recent trial on unguided treatment which included a more comprehensive treatment programme and found a moderate effect size of d = .64 ( Meyer et al., 2009). However, while support increase effects it is associated with more costs ( Palmqvist, Carlbring, & Andersson, 2007). As a correlation of rho = .75 has been observed between the amount of therapist contact in minutes and effect size ( Palmqvist et al., 2007), the input from the therapist in the form of individualized e-mails deserve a more careful test. While the literature on pure e-mail therapy is limited, it is a form of therapy that has been practiced on the Internet for more than 10 years ( Murphy & Mitchell, 1998) and sold by private practitioners ( Manhal-Baugus, 2001). There are no published controlled trials on pure e-mail therapy for depression. However, the border between purely e-mailed based therapy and guided self-help programmes in which individualized therapist input is added is not sharp, and for example one programme called “Interapy” for symptoms of post-traumatic stress ( Lange et al., 2003) adds more therapist time than in most guided self-help programmes. Interapy has also been tested for depression with a promising outcome ( Ruwaard et al., 2009). There is also an emerging literature suggesting that factors such as therapeutic alliance can develop in Internet treatments ( Knaevelsrud & Maercker, 2007) and in e-mail therapy ( Cook & Doyle, 2002). In sum, the literature on the role of the therapist in guided self-help and the lack of robust studies on e-mail therapy suggests that it is not known how effective the individualization inherent in e-mail therapy is. Moreover, it is not known if e-mail therapy yields different outcomes than guided self-help in which minor guidance is given.
The aim of this study was to study the effects of structured guided self-help for depression with minimal therapist contact (Andersson et al., 2005) and e-mail therapy with a strong component of individualization. A waiting-list control group was also included against which the two treatments were compared. The two active treatments were both based on CBT, which in this case involved both components of behavioural activation (Martell, Addis, & Jacobson, 2001), and cognitive therapy as initially described by Beck (Beck, Rush, Shaw, & Emery, 1979).
