Interpretation of observational studies: the good, the bad and the sensational

The increasing cost of research, and pressure on academics to raise their profile, may be contributing factors in the large number of observational studies reported in sensational ways by mass media. Headlines blaming individual foods or nutrients for chronic diseases or, in contrast, implying that eating a particular food could prolong life or drastically cut disease risk, seem all too common yet could be misleading to the consumer thanks to reporting of relative, rather than absolute, risk.In addition, due to the popularity of ‘multiple hit’ statistical analyses or using observational datasets to answer research questions for which they were not designed, there is a risk that statistically significant findings may be spurious or not clinically relevant.A worrying consequence is that observational studies could drive future changes to dietary guidelines or influence research plans, ending up amplified in meta-analyses. Some academic commentators have noted that randomised controlled trials set up to test the conclusions from prospective cohort studies rarely corroborate the findings and can find opposite results e.g. antioxidant supplements and cancer.Perhaps the time has come to establish what the role of observational studies should be in the armoury of nutrition evidence, as well as discussing how the nutrition community can improve the methodology and planning of observational studies in future.Monday, 18 January, 2021 -08:15 to12:15United Kingdom
Source: The Nutrition Society - Category: Nutrition Authors: Source Type: news