IJCS | Volume 32, Nº2, March/April 2019

164 Lei & Bin Differences of risk factors in young AMI patients Int J Cardiovasc Sci. 2019;32(2)163-176 Review Article risk factors for AMI in young patients. 7,8 This article aimed to assess the differences in risk factors and clinical characteristics between young and older AMI patients. Methods Data sources and search strategies A search was made in the database of Cochrane Library, Pubmed, BioMed Central and Embase, since their establishment to December 2016. An experienced searcher used the key words: risk factors, clinical characteristics, young people and acute myocardial infarction, with the Boolean operators AND and OR. We searched for comparative studies of risk factors and clinical characteristics in myocardial infarction between young and older patients. The search was limited to observational studies on humans of the randomized controlled trial (RCT) type. For the meta-analysis, we only used articles published in English. Study selection and extraction criterion Most studies used an age cutoff of 40 to 45 years to define young patients diagnosed with AMI, thus we chose patients aged 45 or less as the limit for young AMI patients, while patients aged older than 45 years were defined as older AMI patients. We reviewed the list of identified articles and extracted data from the selected ones; subsequently we selected studies with abstracts suggesting they were relevant. Studies were eligible if: (1) the study design was a cohort or case-control study and all the studies were RCTs; (2) the study compared young AMI patients with older AMI patients; (3) the study reported risk factors or clinical characteristics of young AMI patients, including any ethnicities and nationalities. Initial abstract screening excluded non- relevant and non-original studies, then full-text review excluded ineligible studies as follows: (1) studies without comparison between young and older AMI patients; (2) age-definition for the young AMI patients was older than 45 years or less than 44 years; (3) studies with abstract only or studies without full-text available; (4) studies lacking complete important information or those with no reply from the contact author; (5) smoking patients were defined as current smokers, and former smokers were excluded. For each study, we recorded the following information: first author, year of publication, number of cases of young and older AMI patients, risk factors and clinical characteristics. Risk factors of AMI were: smoking, Hypertension, family history of CAD, obesity, hyperlipidemia, diabetes mellitus, alcohol consumption. We defined Hyperlipidemia as a condition with elevated serum lipid levels, including high levels of total cholesterol (TC) or elevated levels of low-density lipoprotein (LDL), or high triglycerides (TG). To better assess the effect of serum lipids on myocardial infarction in young people, we also compared high-density lipoprotein (HDL) levels in young and older AMI patients. Clinical characteristics were: chest pain, left ventricular ejection fraction (LVEF) value (%), all-cause mortality and outcome of coronary angiography (CA). Literature quality assessment We assessed the literature quality using the standard bias risk assessment of the Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0,9 of which scale consists of: random sequence generation, allocation concealment, blinding of participants and personnel, complement of outcome data, incomplete outcome data, other bias resource. The risk bias of each study uses “High risk”, “Low risk” and “Unclear risk” for each scale. Statistical analysis We used the Review Manager 5.3.0 software for comprehensive meta-analysis. We used the X 2 test and I 2 statistics (ranging from 0% to 100%) to estimate the percentage of total variation across studies. When p ≥ 0.1 and I 2 valued 50% or less, the data showed low heterogeneity andwe used the fixed-effect model to pool results across studies. When p < 0.1 and I 2 values were higher than 50%, the data showed high heterogeneity and the random-effect model was used to pool the results from studies, and a subgroup data analysis was also performed. When an extremely high heterogeneity influenced the determination of its resource, the description analysis was used as presentation. For each risk factor compared between young and older AMI patients, we calculated the adjusted odds ratio (OR) and corresponding 95% confidence interval (95%CI) in each study. Funnel plots were used to estimate publication bias. All P values were two-tailed, and a p value < 0.5 was considered significant.