[Body mass index and all-cause mortality in Asian adults: a meta-analysis]

To perform a systematic review between all-cause mortality and body mass index (BMI) in Asian adults.

Relevant prospective studies that reported the relative risks (RRs) of all-cause mortality for community-based adults in Asia were identified by a literature search.PubMed and CNKI electronic databases were searched from inception through September 30, 2014, with language restrictions of English and Chinese. Data were extracted by 1 reviewer and then reviewed by 3 independent reviewers. The overall effect of varied levels of BMI and all-cause mortality were then pooled and analyzed.Potential sources of heterogeneity were detected by stratification analyses and sensitivity analyses. Publication bias was detected by funnel plot, Egger's test and Begg's test.

Seventeen studies met the inclusion criteria; these studies included 1 769 369 individuals with 104 888 deaths. Random-effects summary all-cause mortality RRs was calculated. With the use of a BMI (in kg/m²) of 18.5~22.9 as the reference, the summary RRs were 1.39(95% CI: 1.31~1.47) for BMI less than 18.5 kg/m² , 0.88 (95% CI: 0.85~0.92) for BMI of 23.0~29.9 kg/m² and 1.14 (95% CI: 1.05~1.23) for BMI more than 30.0 kg/m². The RRs tended to be higher when weight and height were self-reported rather than measured. The RRs were higher when papers were published before the year 2005 rather than after the year 2006. Also, the RRs were higher when the quality scores were higher. Potential sources of heterogeneity were gender, the method of obtaining weight and height, geography, publication year and quality scores. There was no publication bias (P>0.05) in this meta-analysis.

There was an increased risk of all-cause mortality for those both at the lower and higher level of BMI in Asian adults.