Evidence mapping: methodological foundations and application to epidemiologic research on sugar sweetened beverages and health

Article type
Authors
Althuis MD1, Weed DL2
1EpiContext, USA
2DLW Consulting Services, USA
Abstract
Background: Evidence maps are a new methodology that systematically characterizes the range of research activity on broad topics and are used to guide research priorities, systematic reviews, and meta-analyses. This methodology is underused in epidemiology and has considerable potential to increase the efficiency of research efforts.

Objectives: We demonstrate the usefulness of evidence mapping as a tool for organizing epidemiological studies, using as an example research on sugar sweetened beverages (SSB) and health outcomes: obesity, type 2 diabetes (T2D), metabolic syndrome (MetSyn) and coronary heart disease (CHD)/stroke.

Methods: We performed a search of Pubmed, Scopus, and Cochrane databases and a hand search of references. Studies selected were published reviews and longitudinal studies (intervention and cohort; January 1, 1966–October 31, 2012).

Results: We identified and mapped 77 studies (18 review, 59 primary research). Most research focused on obesity (N = 47), with the numbers of studies reducing to 6–11 when categorized into groups by age and study type (Fig. 1).The number of cohort studies that assessed SSB and remaining health outcomes were T2D (N = 9), MetSyn (N = 4), and CHD/Stroke (N = 4). For all outcomes, more than 30% (N = 18) of the primary research studies we identified were not referenced in published reviews. We found considerable variability among primary research studies of SSB and the four health outcomes in terms of designs, definitions of SSB, and definitions of outcomes. For example, we counted 14 different definitions of weight/obesity in 29 cohort studies, with no more than 6 studies reporting use of the same outcome measure.

Conclusions: This map showcases the complexity of research on this topic. Establishing standards in the study of SSB and health would facilitate interpretation across research studies, thereby increasing utility of systematic reviews/meta-analyses and ultimately efficiency of research efforts. Rapid publication of new data suggests the need for caution when reading reviews and regular updates.