To our knowledge, this is the first study to assess the number of articles published on eating disorders across cultures. The number has increased exponentially between 1970 and 2011, indicating considerable and increasing interest and awareness in this area. The most recent two-year time bracket (2010–11) showed 219 articles had been published, nearly double that in the five-year bracket of 1990–94. This is heartening, given that from 1996 to 2010 alone, 4,162 articles were published by six specialized eating disorders journals .
Of all the cultural groups analysed, the African group (including African Americans) has been the best studied. This group was one of the first to be investigated, and currently one-fifth of cross-cultural articles about eating disorders have been on African cultural and ethnic groups. Interestingly, the proportion of publications on this group has decreased slightly over the last decade. However, studies of African groups still form a substantially greater proportion than do studies of all the other groups investigated. The second-most studied was the Latin American and Hispanic group. All the other ethnic groups have had very few articles published.
It is not clear why the South European group should have been so poorly studied, particularly considering Spain was included in this group and that the Latin American/Hispanic (and thus predominantly Spanish-speaking) group has been comparatively well represented. It may reflect a publication and indexing bias towards US articles, which in turn have a particular interest in both African American and Hispanic groups. The Pacific Islands and South Asia have also been poorly represented. The number of articles published on Middle Eastern and East European groups, while small, has steadily increased over time. For East and South-East Asians, the percentage of manuscripts devoted to them has plateaued. The limited epidemiological data suggests that the prevalence of eating disorders in non-Western groups is on the rise [2, 9]; however, this may merely reflect growing awareness and identification of eating disorders in these populations. Nevertheless, it is concerning that there are so few studies of many of these groups and that the number and proportion of studies is not necessarily increasing over time.
With regard to the types of papers published in cross-cultural eating disorders research, most are empirical or comparative studies. It was expected that there would be few controlled trials, as such research is very difficult to undertake in eating disorders in general. In earlier decades, empirical studies formed the vast majority of cross-cultural publications and this remains the case in 2010–11, although there appears to be an increase in validation studies. This latter finding should be interpreted cautiously, as the number of publications classified as validation or psychometric studies is still small.
The limitations of this analysis are that we used Medline and PsycInfo, which are popular databases but mostly index journal articles which are in English or have English language abstracts [10, 11]. This potentially makes non-English language research “invisible” to academics in English-language settings, who are thus not aware of and/or not accessing such material. Our analyses show a small but steady growth in the number of articles indexed by Medline and PsycInfo which are published in languages other than English. This may not be a true reflection of publication rates of non-English language articles, and non-English databases should be searched in future studies. For example, Chisuwa and O’Dea  have searched Japanese databases to locate eating disorder articles in Japanese journals, citing 11 papers, including publications on the prevalence of eating disorders in Japan, treatment infrastructure and body image. Another limitation to our study is that the subject headings used to extract citations from Medline and PsycInfo did not yield mutually exclusive sets of articles. For example, Egyptian populations were indexed as both Middle Eastern and African. Studies which cover more than one ethnic group would also be indexed and counted under individual ethnicgroups. Finally, some articles were indexed under more than one publication type.