Main

Ethnic and racial variation in disease prevalence and health care use is the focus of much current research. The burgeoning literature has been accompanied by a diverse range of concepts and terms, often confusing and inappropriate, to describe the population under study.1 This is not surprising as race and ethnicity are complex, multidimensional concepts changing with time.1,2,3 Explicit definitions of concepts and terms are clearly essential to enable the reader to understand reported research, and to permit comparisons between studies, particularly internationally. Yet, the number of studies which provide such definitions is few; less than 15% were found to do so in two recent reviews.4,5

Debate on the concepts and terminology of ethnicity and health research is vigorous in the USA and emerging in the UK. The key issues include the definitions and measurement of race and ethnicity; the interpretation of words such as Asian, South Asian, Latino, Hispanic, Black, and White; and the pitfalls of using superficial labels as a description of study populations.1,4,5,6,7

Editors are responsible for ensuring scientific rigour and high quality writing in their journals. To help inform the current debate on terminology in ethnicity and health research, we undertook a survey of journal editors in 1994 (74% response rate) to ascertain editors' current practice and views and to stimulate discussion on the issue of terminology in ethnicity and health research.6 Of the 38 journals surveyed, only one (4%) had a policy on terminology. However, 23 (82%) editors thought the issue was an important one, and 16 (57%) thought it was worthy of discussion by the editorial board. Our second survey (Rankin J, Bhopal R, unpublished), two years later, in 1996 (55% response rate) examined whether editors had altered their views or practice. Two (11%) editors reported a change in editorial policy, although neither had a written policy, and one (5%) reported a change in the journal's instructions to authors. In fact, 14 (74%) editors now said they did not think a written policy was required (Rankin J, Bhopal R, unpublished observations). In a survey of 29 editors of US public health and medical journals, Bennett and Bhopal found most journals did not have a policy on terminology, and showed modest commitment to changing editorial policy or instructions to authors.7 The need for authoritative guidance was raised in both UK surveys. Editors who were keen to continue the debate suggested discussion in journals and at conferences. There was some desire for a consensus policy which could be adopted, though this was tempered by the wish of editors to retain control of journal policy.

Until the ongoing debate yields workable solutions, researchers should do their best to ensure that appropriate and consistent concepts and terms are used to describe the population under their study. Current advice is summarised in Table 1. We emphasise three principles which we commend to BDJ readers and writers:

  1. 1

    The emergent concept of race emphasises its social origins rather than its biological basis

    The concept of race as reflecting genetically different human populations is scientifically weak and should be avoided.3,8 The emergent concept of race emphasises its social origins rather than its biological basis. Race provides a way of defining, for social purposes, populations which look different and have different ancestral roots.9 The term race should be used with caution for its history is one of misuse and injustice.8,10

  2. 2

    The concept of ethnicity refers to the social rouping(s) people belong to because of their culture, which includes language, religion, dietary and marital customs and other factors which relate to ancestry

    Table 2 Table 2
    Full size table

    The concept of ethnicity refers to the social grouping(s) people belong to because of their culture, which includes language, religion, dietary and marital customs and other factors which relate to ancestry.3 Ethnicity is fluid and changeable. Ethnicity and race are overlapping but different concepts which should not be used synonymously, but often are. Ethnic labels, as in the 1991 Census (Table 2), are no more than a first step to defining a person's ethnicity (see Smaje for discussion11).3,11

  3. 3

    Labels such as White, Asian, Latino, Afro-Caribbean, black, need to be recognised as inaccurate and crude shorthand for potentially important information about a person's ethnicity.1,12 The need for simplicity should be weighed against the dangers of stereotyping and inaccuracy. As a minimum, writers should define these terms. Better, they should provide a description of the population they are referring to.13,14 For example, the label 'South Asian' should not be used if the population referred to is a Bangladeshi one. The tendency to lump together diverse populations is harmful. For example, Bangladeshi men have an extremely high prevalence of smoking, a fact lost by studies of 'South Asians'.

Table 1 Table 1
Full size table

Readers and writers of the BDJ need to be tuned into the ongoing debate. Editors need to be involved in a leadership role. Dental problems, in common with most diseases, vary by ethnicity and race (however defined and classified), with implications for health care delivery and preventive care. Dental practitioners and researchers need to be aware of such variations, and will continue to research the factors which underlie them.15,16 Accurate use of concepts and words is an essential first step to good research, to improving the health of ethnic minorities and narrowing inequities.