The researchers, led by Douglas Weed of Emory University’s Rollins School of Public Health, wrote that controversial topics with important implications for public health, like the effect of sugar-sweetened beverages (SSBs) on health, need good quality, rigorous scientific reviews to ensure that scientific and public health decisions are built on a solid foundation.
However, they found that reviews published over the past decade linking SSB consumption with obesity, metabolic syndrome, type-2 diabetes and coronary heart disease were of moderate low-quality overall.
Using a system called AMSTAR (assessment of multiple systematic reviews), they found: “Less than one-third of published reviews reported a comprehensive literature search, listed included and excluded studies, or used duplicate study selection and data abstraction.”
The research, which received funding from the Coca-Cola Company, did not seek to examine the conclusions of the scientific reviews about the possible health effects of sugary drink consumption.
The authors concluded by urging researchers to insist on methodologically systematic reviews and urging educators to train students in the methodology of systematic reviews.
Distorted results?
In an accompanying editorial, authors Vasanti Malik and Frank Hu from the Harvard School of Public Health and Harvard Medical School said that the view put forth in the paper of the state of the literature on SSBs and health outcomes was distorted by lumping systematic reviews, which received high scores under the AMSTAR system, together with non-systematic reviews, which received low scores.
In particular, they criticized the authors’ use of AMSTAR for assessing narrative reviews, which summarize the literature on a given topic without providing a detailed meta-analysis, and often do not include a breakdown of how a literature search was carried out.
“By applying the AMSTAR instrument equally to narrative and systematic reviews, the authors underestimated the quality of the reviews of SSBs and health, leading to a spurious conclusion,” wrote Malik and Hu.
They added: “A systematic review of reviews, like the individual reviews themselves, should be of sufficiently high quality and methodologic rigor so that reliable conclusions can be drawn.”
Source: American Journal of Clinical Nutrition
2011; 94:1340-7
"Quality of reviews on sugar-sweetened beverages and health outcomes: a systematic review"
Authors: Douglas L. Weed, Michelle D. Althuis, and Pamela J. Mink