Diet in cardiovascular diseases: an overview of umbrella reviews and overviews of reviews

Article type
Authors
Posadzki P1
1Kleijnen Systemati Reviews, Ltd, Escrick, Yorkshire, United Kingdom
Abstract
"Background: Cardiovascular diseases (CVD) are leading cause of morbidity and mortality worldwide. The role of unhealthy diet being a significant risk factor for CVD has extensively been studied. Umbrella reviews (URs) and/or overviews of reviews (OoR) are tertiary research evaluating a breadth of evidence for a particular health topic.
Aim: This overview aims to synthesise evidence from umbrella reviews (URs) and overviews of reviews (OoR) evaluating the role of dietary interventions in CVD.
Methods: PubMed, Cochrane Central and Epistemonikos were searched from their inceptions to January 18th, 2024, without language restrictions. Only URs and OoRs of both observational and interventional studies of dietary interventions (also in the context of prevention) in CVD were included. Non-systematic URs/OoRs, clinical practice guidelines (CPGs), or those evaluating e.g., anthropometrics only were excluded. No methodological/risk of bias evaluations were undertaken since a tool for umbrella reviews is yet to be developed. Data were synthesized narratively.
Results: 35 URs/OoRs met the eligibility criteria. The URs were published between 2012 and 2023. Those URs/OoR pertained to Dietary Approaches to Stop Hypertension (DASH) (n=1), chocolate (n=1), coffee (n=1), [abstract truncated]. The URs predominantly included SRs that relied on mixed studies (cohort and randomised controlled trials) (n=23, 65.7%), cohort studies only (n=8, 22.8%), and randomised controlled trials only (RCTs) (n=4, 11.4%). Twenty-six URs (n=26, 74.2%) found positive associations between various diets/nutrients and improved CVD outcomes, three (8.5%) found negative associations between milk, diary and egg consumption and CVD, and six (17.1%) found unequivocal associations. Risk of bias in SRs (AMSTAR, ROBIS, Oxman score) was predominantly moderate to high (n=14, 40%), low (n=5, 14.2%), or inestimable/unclear (n=16, 45.7%). Certainty of evidence (CoE) ranged from very low to low (n=12, 34.2%), moderate (n=3, 8.5%), or it was inestimable/unclear (n=20, 57.1%)
Conclusions: There is an abundance of URs/OoRs evaluating the effects of diet/nutritional interventions (some of them overlapping especially in mixed diets/dietary patterns) in CVD. Amongst the included SRs, ROB was moderate or high, and the CoE was predominantly very low to low indicating the existing uncertainties. "