Effect sizes of associations between neuroimaging measures and affective symptoms: A meta-analysis

Background The utility of brain-based biomarkers for psychiatric disorders hinges among other factors on their ability to explain a significant portion of the phenotypic variance. In particular, many small scale studies have been unable to arbitrate whether structural or functional magnetic resonance imaging has potential to be a biological marker for these disorders. Methods This study conducted a meta-analysis to examine the relationship between study power and published effect sizes for the relationship between affective symptoms and structural or functional magnetic resonance imaging measures. The current analyses are based on 821 brain-affective symptom association effect sizes derived from 120 publications, which employed a univariate region-of-interest approach. Results For self-assessed affective symptoms published brain imaging measures accounted for on average 8% (confidence interval: 1.6%-23%) of between-subject variation. This average effect size was based mostly on studies with small sample sizes, which have likely led to inflation of these effect size estimates. Conclusions These findings support the conclusion that brain imaging measures currently account for a smaller proportion of the interindividual variance in affective symptoms than has been previously reported. The current findings support the need for both large-sample clinical studies and new statistical and theoretical models to more robustly capture systematic variance of brain-affective symptom relationships.

Automated summary

The meta-analysis found that current brain imaging measures account for a smaller proportion of interindividual variance in affective symptoms than previously reported. The results suggest the need for large-sample clinical studies and new statistical and theoretical models to better capture systematic variance in brain-affective symptom relationships.