Search strategy

The electronic databases for literature search will include PubMed, Web of Science, and Scopus. For search of the eligible studies, the following keywords and Medical Subject Heading will be used:

#1 “epicardial adipose tissue” OR “epicardial fat” OR “pericardial adipose tissue” OR “pericardial fat” OR “cardiac adipose tissue” OR “cardiac fat” OR “subepicardial adipose tissue” OR “subepicardial fat” OR “heart fat” OR “heart adipose tissue”

#2 “left ventricular” OR “left ventricle” OR “systolic function” OR “systolic dysfunction” OR “diastolic function” OR “diastolic dysfunction” OR “left atrium” OR “left atrial”

#3 “cardiovascular” OR “hospitalization” OR “heart failure” OR “myocardial infarction” OR “unstable angina” OR “prognosis”

#4 #1 AND #2 (primary outcome)

#5 #1 AND #3 (secondary outcome)

Literature search will also be conducted by manual screening of reference lists of relevant reviews and retrieved articles. Two researchers (HF and TG) will independently perform the literature search. Disagreements will be resolved by consensus. The literature search will be repeated before completing the data extraction, and potential new studies published during the work process will be added to the result. Study selection will be conducted in a PRISMA-compliant flow chart (Fig 1). Only articles published in the English language will be included.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA flow diagram.

https://doi.org/10.1371/journal.pone.0283482.g001

Study design

Observational studies will be included. Case-control studies will be excluded.

Selection criteria

Inclusion criteria for this meta-analysis included: (1) included adult (>18 years) patients; (2) quantified epicardial fat volume using CT or MRI; (3) measured cardiac structure and function on echocardiography, CT or MRI or assessed cardiovascular outcomes; and (4) described the relation of epicardial fat volume with cardiac structure and function or cardiovascular outcomes.

Outcomes

The primary outcome will be cardiac structure and function. The secondary outcome will be cardiovascular events including death from cardiovascular causes, hospitalization for HF, non-fatal myocardial infarction, and unstable angina. In the measures of cardiac structure, LV mass and left atrial volume will be extracted. In the measures of LV systolic function, LV ejection fraction and early systolic mitral annular velocity (s’) will be extracted. In the measures of LV diastolic function, early diastolic mitral annular velocity (e’) and the ratio of early diastolic mitral inflow to annular velocities (E/e’) will be extracted given the linear relationship with LV diastolic dysfunction grade.

Data extraction

Two reviewers (HF and TG) will independently extract relevant data from retrieved studies, including author, study design, study time, country, number of participants, baseline characteristics (age, sex, body mass index, epicardial fat volume, used modalities for assessing epicardial fat volume and cardiac structure and function, etc), clinical outcomes (cardiac structure and function and cardiovascular events), and information on the methodological quality (selection of cohorts, assessment of outcome, etc). Disagreements will be resolved by consensus. We will contact the corresponding author of eligible studies when insufficient information is available to perform our meta-analysis.

Quality assessment

The quality of observational studies will be evaluated by Newcastle-Ottawa Scale tool (http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp).

Statistical analysis

To assess the association of pericardial fat with cardiac structure and function, correlation coefficients between pericardial fat volume and indices of cardiac structure and function will be synthesized. To assess the association of epicardial fat with cardiovascular outcomes, hazard ratios adjusted by variables of age and sex at least will be synthesized. For each outcome, heterogeneity will be assessed using the Cochran’s Q and I² statistic; for the Cochran’s Q and I² statistic, a p value of <0.1 and I²>50%, will be considered significant, respectively. When there is significant heterogeneity, the data will be pooled using a random-effects model, otherwise a fixed-effects model will be used. When there are more than 10 studies included, publication bias will be assessed graphically using a funnel plot and mathematically using Egger test. For these analyses, Comprehensive Meta Analysis Software version 2 (Biostat, Englewood, NJ, USA) and STATA 16 software (Stata Corp LP, TX, USA) will be used.

For outcomes with insufficient data, a systematic narrative synthesis will be provided with information presented in the text and tables to summarize and explain the characteristics and findings of the included studies. The narrative synthesis will explore the relationship and findings both within and between the included studies, in compliance with the guidance from the Centre for Reviews and Dissemination [11].

Sensitivity analysis

Subgroup analysis stratified by study design (prospective or retrospective), modalities for quantifying epicardial fat volume (CT or MRI), and baseline patient characteristics (patients with cardiovascular diseases or general population, obese or not, elderly [≥65 years] or not) will be performed. Meta-regression will be used to determine whether the association of increased epicardial fat with outcomes will be confounded by baseline clinical characteristics (age and body mass index).

Ethical issues

This meta-analysis is a literature study. Ethical approval is not required because this meta-analysis will not involve any subject directly.