Mindfulness and Depression. Although mindfulness has been proposed to have benefits for a number of psychological disorders (Baer, 2003; Grossman, 2004), the present discussion will focus specifically upon depression, as the meditators' postintervention decrease in depressive symptomatology, relative to the nonmeditators', represents the most significant change in the affective measures employed in the present study. Furthermore, two measures of improved executive function were found to at least partially mediate the relationship between mindfulness training and reduced depressive symptomatology. Specifically, there was a strong trend toward enhanced executive function, namely decreased RTs on the neutral condition of the IST, mediating the effect of mindfulness training on reduced depressive symptomatology. This result adds credence to the notion that executive function may be one of the mechanisms underlying mindfulness (Baer, 2003; Bishop et al., 2004).
This inverse relationship between executive function and depression has been reported in a large body of previous research (Bishop, 2002; Borkovec, 2002; Brown and Ryan, 2003; Carlson, Speca, Patel & Goodey, 2003; Heidenrich & Michalak, 2003; Kabat-Zinn et al, 1992; Lowenstein, 2002; Ma & Teasdale, 2004; Mason & Hargreaves, 2001; Miller, Fletcher & Kabat-Zinn, 1995; Shapiro, Schwartz & Bonner, 1998; Teasdale, 1999; Teasdale, Segal & Williams, 1995; Teasdale, Segal & Williams, 2003; Williams, Teasdale, Segal & Soulsby, 2000). However, these studies have tended to focus on clinically depressed populations. More research is required in order to determine whether the conclusions of these studies hold when examining the effects of mild depressive symptomatology. However, the present results do appear to demonstrate a relationship between decreased depression and improved executive function, in a sample marked by variations within the normal range of depressed mood, consistent with the few previous studies which have investigated this relationship (Berndt & Berndt, 1980; Farrin, Hull, Unwin, Wykes & David, 2003; Slife & Weaver, 1992). These have found that depression tends to impair executive function, although there is debate as to the extent and specific nature of this dysfunction.
Berndt & Berndt (1980) report one such area of depression-related dysfunction to be increased RTs on complex psychomotor tasks involving memory and executive function (such as digit symbol tests and Neckar Cube reversals). A similar pattern has been observed in cases of severe depression (Sobin & Sackheim, 1997). This suggests that the decrease in the meditators' postintervention RTs observed here may reflect enhanced executive function resulting from decreased depressive symptomatology. Interestingly, the observed increase in mindfulness amongst the meditators was found to be mediated both by reduced RTs on the neutral condition of the IST and enhanced working memory capacity. This is consistent with previous research that suggests working memory and executive function refer to the same construct (Baddeley, 1986; Ruff & Rothbart, 1996), and supports the notion that enhanced executive function may at least partially mediate mindfulness (Bishop et al., 2004). The fact that this may be enhanced by training provides evidence for the notion that executive function is a limited-capacity system (Posner, 1995). More importantly, however, this finding suggests that enhanced executive function may generate improvements in affective function. This is consistent with a number of accounts in the literature, which have suggested that this relationship is mediated by self-regulation (Mishel & Ayduk, 2002; Rueda, Posner & Rothbart, 2004; Ruff & Rothbart, 1996). Mindfulness training thus appears to benefit self-regulation.
Slife & Weaver (1992) suggest that executive function represents a metacognitive skill, and that deficient metacognitive processing, resulting from depression, may generate specific cognitive deficits such as rumination and working memory dysfunction. The results of the present study appear to support this. Enhanced executive function, characterised by decreased RTs and enhanced working memory, may lead to increased mindfulness and concomitantly improved affect. However, it should be noted that the present results do not support the inverse relationship between mindfulness and rumination which has been proposed in previous studies.
It is interesting that the observed improvement in depression levels amongst the meditators occurred in the absence of a significant improvement in levels of rumination. This is difficult to reconcile with past research which has suggested a link between the two (Treynor, Gonzales & Nolen-Hoeksema, 2003; Watkins & Teasdale, 2001; Roemer & Orsillo, 2002). Two explanations are offered here. It is possible that mindfulness operates more upon depressive symptomatology than ruminative patterns. More research is needed in this area before any solid claims can be made. Alternatively, this may reflect issues of power. It should be noted that there was a weak trend toward a significant time by group interaction effect on the RRS. Although the effect size for this interaction was small, it is possible that a replication of this study, with a larger sample size, may find a significant postintervention reduction in rumination.
The Differential Relationship between Measures of Affect. The apparently differential relationship between the measures of affect assessed in the present study may be more generally explained by models which suggest that positive and negative affect may vary on separate dimensions, rather than a single bipolar dimension (Diener, 1999; Diener & Emmons, 1985; Watson, Wiese, Vaidya & Tellegen, 1999). Such models offer an explanation for the present findings that changes in depression, anxiety, positive and negative affect may vary independently of one another. Again, researchers may wish to consider this when designing future studies aimed at elucidating the relationship between mindfulness and improved affect.