More than the sum of its parts: meta-analysis and its potential to discover sources of heterogeneity in psychosomatic medicine.

Meta-analyses may contribute to more reliable knowledge about the existence of certain relations in the area of psychosomatic research. Surprisingly, the increasing popularity of meta-analysis is not reflected in the number of meta-analyses of observational studies published in Psychosomatic Medicine. This may be due to the specific difficulties that apply to meta-analyses of observational research. The aim of this paper is to provide a nontechnical overview of the principles of meta-analysis applied to observational research. We will highlight general principles of meta-analysis and discuss the major threats to its validity, with an emphasis on its specific merits and pitfalls for psychosomatic research, using several examples. We conclude that meta-analysis is a relatively simple technique, leaving little reason for not routinely applying it when performing a systematic review. An adequately conducted meta-analysis may not only provide a summary estimate of a certain association, but it has additional value in discovering relevant confounders, mediators, and moderators, as well as identifying areas of research that require more attention.