LAUSR

BACKGROUND Vascular mechanisms may play a role in depression. The aim of this review is to summarise the evidence on alterations in cerebral haemodynamics in depression. METHODS MEDLINE (1946- present), Embase (1947-present), Web of Science (1970-present), PsycINFO (1984-present), CINAHL (1976-present) and CENTRAL were searched using a predefined search strategy. A meta-analysis was conducted in four groups: 1) global cerebral blood flow (CBF) in ml/min/100g, 2) CBF velocity (CBFv) in cm/s (maximum flow of left middle cerebral artery, 3) combined CBF and CBFv, 4) Ratio of uptake of Tc 99m HMPAO (region of interest compared to whole brain). Data are presented as mean difference or standardised mean difference and 95% confidence interval (95% CI). A narrative synthesis of the remaining studies was performed. RESULTS 87 studies were included. CBF was significantly reduced in depressed patients compared to HC [15 studies, 538 patients, 416 HC, MD: -2.24 (95% CI -4.12, -0.36), p=0.02, I2=64%]. There were no statistically significant differences in other parameters. The narrative synthesis revealed variable changes in CBF in depressed patients, particularly affecting the anterior cingulate and prefrontal cortices. LIMITATIONS There were various sources of heterogeneity including the severity of depression, use of antidepressant medication, imaging modality used and reporting of outcomes. All of these factors made direct comparisons between studies difficult. CONCLUSIONS The reduction in CBF in depressed patients compared to HCs may indicate a role for assessment and CBF altering interventions in high-risk groups. However, results were inconsistent across studies, warranting further work to investigate specific subgroups.