Ires Ghielen

112 Chapter 6 risk of bias were analyzed. In these post-hoc analyses, the effect sizes decreased to small ( g = 0.31, g = 0.33, g = 0.36, g = 0.39, g = 0.42, respectively). No evidence was found for publication bias. The small to moderate main effect sizes suggest that CBT and MBT are beneficial in reducing psychological distress, but only to a certain extent. Biological approaches, e.g. pharmacotherapy, showed a reduction of depressive symptoms in MS patients with an effect size of 0.63 (standardized mean difference) [22]. According to the review and meta-analysis by Fiest et al. [22], current research is insufficient to determine the effectiveness of pharmacotherapy for anxiety in MS as no controlled studies were found. In PD, the meta-analysis of Bomasang and colleagues [51] on antidepressant medication showed an effect size of 0.54 in reducing depressive symptoms. The effect of pharmacotherapy on reducing symptoms of anxiety in PD patients has insufficiently been studied. Although the effect sizes of pharmacotherapy on depressive symptoms appear to be larger than those of psychological treatment, regarding anxiety and global mental health the effect is not yet investigated properly. One can imagine that pharmacological interventions show larger effect sizes compared to psychotherapeutic interventions, since the latter requires cognitive abilities to learn and apply the methods that are taught. Although patients with dementia were excluded in most studies, it is possible that these populations have reduced cognitive abilities as a result of the neurodegenerative process, and are therefore unable to optimally benefit from CBT and MBT. It is also argued that a combination of psychotherapy and pharmacotherapy might be most beneficial, at least for outpatients with chronic forms of depression [52, 53] and panic disorder [53]. In adults with an anxiety or depressive disorder without neurological comorbidity, a meta-analysis of Cuijpers and colleagues [54] showed that CBT is probably effective. Although effect sizes were larger (around g = 0.80) compared to our results, the quality of the included studies was low and publication bias was present. Large effect sizes were also found for MBTs in the treatment of anxiety and depressive symptoms in participants without neurological comorbidity [55]. Here, no publication bias was present but study quality was again unsatisfactory. CBT and MBT appear to be more effective in patients without compared to patients with neurodegenerative disorders. However, the methodological quality is insufficient to draw definite conclusions. MS patients seem to benefit more from CBT and MBT than PD patients as is represented by the significant subgroup difference in effect size regarding disease type. However, considerable heterogeneity was present in both subgroups and all therapies described here were adapted to the respective study sample. The MS