Meta-analysis in Research on Nutrition

摘要

To the Editor Dr Barnard and colleagues~(1) questioned the validity of meta-analyses in nutrition research, highlighting the industry-funded meta-analysis by Chowdhury et al in 2014 that failed to support the diet-heart hypothesis, that excess saturated fat intake causes cardiovascular disease. Their argument was that nutritional epidemiological studies are complex interventions that do not lend themselves to aggregation, compared, for example, with drug trials. However, meta-analyses are useful summaries, offering the reader an overview of the available literature, while addressing the issue of type 2 error by pooling associations or effect measures of individual studies or trials.