The Grocery Gap: Who Has Access to Healthy - The Food Trust

PolicyLinkThe Food TrustMethodsBetween May and July 2009, PolicyLink andThe Food Trust created a comprehensivebibliography of studies related to food accessand/or food access and health across theUnited States. Unlike previous review studies,which typically only include published work, wesought to include “grey literature,” or studies,reports, and analyses that are not published inpeer-reviewed journals. Public health agencies,community-based organizations, and policygroups frequently conduct primary data analysesof retail food access to inform their activities,but generally do not take the additional steps tosubmit their studies to journals for publication.We used the following search methodsto compile the bibliography:• Sent requests for information to relevantlistservs, e.g., COMFOOD, the NationalNeighborhood Indicators Partnership(NNIP), American Evaluation Association(EVALTALK), and agency email lists;• Wrote to 80 food policy councils acrossthe country that are listed on the NorthAmerican Food Policy Council websiteand the Community Food SecurityCoalition Food Policy Council Database;• Contacted several foundations andleaders working in the food access field;• Searched PubMed and other librarydatabases related to the fields ofplanning, community development, andgeography to identify formally publishedwork related to urban and rural foodaccess and health implications; and• Reviewed reference lists of included studies.In order to be included, each study neededto meet the following criteria:• Related directly or indirectly to identifyingdisparities in access to food retailersor healthy food, and the relationshipbetween food retail and health;• Either included original research onthese topics or reviewed other studies;• Conducted in the United States (while therehave been studies conducted outside ofthe United States, the persistent trend ofresidential segregation by race/ethnicity andincome in this country makes extrapolationfrom these studies of limited value); and• Published during or after 1995 (althoughwe included a few important studies thatwere conducted between 1990 and 1994).One hundred and thirty-two studies were ultimatelyincluded in the database, of 168 articles initiallygathered through the search methods above. Weincluded studies that use random and nonrandomsampling methods and quantitative and qualitativetechniques (such as resident interviews). We alsoincluded studies that examine single communities ofinterest (alone or in comparison to other areas). Weexcluded newsletters, policy statements, and studiesthat focused on methods and measurements. Inone case we found two policy papers, one shorterthan the other, based on the same study and data;they were counted as one study in the database.Of the studies selected for the database, 61 werepublished in peer-reviewed journals, and 71 fellinto the grey literature category. We did notsystematically review the evidence quality (e.g.,sample size, strength of methods used) of each23