The Roma (Gypsies) became one of the peoples of Europe around one thousand years ago, when they first arrived in the Balkans [1,2]. The current size of the European Romani population, around 8 million [2], is equivalent to that of an average European country (Figure 1). While human rights and socio-economic issues related to the Roma are increasingly becoming the focus of political debate and media coverage throughout Europe, their poor health status [[3–6]] is rarely discussed and still awaits the attention of the medical profession.

This review of genetic studies of the Roma was prompted by two recent developments: (i) Studies conducted over the last decade have resulted in the identification of a number of novel single gene disorders and disease-causing mutations. The accumulating data are already sufficient to outline a pattern and draw conclusions about public health policies and future research. (ii) The economic and political changes in Eastern Europe and the wars in former Yugoslavia have led to the west-bound migration of large numbers of Roma [7,8], changing the traditional demographic profile of Gypsy minorities across Europe. A predictable consequence of this new diaspora is that medical practitioners in many countries will encounter Romani patients with previously unknown or very rare disorders. A summary of the available information should facilitate diagnostic investigations and counselling in these affected families and stimulate international collaboration.

Literature searches were performed using the U.S.A National Library of Medicine PubMed/MEDLINE databases for the period 1960 to December 2000. Database searches using the keyword "Gypsies" identified 297 articles whilst the keyword "Gypsy" produced 573 articles. The discrepancy is due mainly to the inclusion of articles about the "gypsy retransposable element" and the "gypsy moth". Searches using the terms "Roma", "Romani" and "Romany" yielded results that were not relevant to the topic (eg. Roma, the capital of Italy) or else incomplete.

The majority of the 297 articles dealt with issues beyond the focus of this review, namely social problems related to the health of the Roma (28.6%), or general medical problems (29.6%). The remainder were reports on genetic research, of which 41 studies (13.8%) were in the field of clinical genetics, 44 (14.8%) were molecular studies of genetic disorders, and 39 (13.1%) covered population genetic research. In the clinical and molecular genetics fields, we have given preference to publications which were not limited to single case descriptions, and dealt with disorders with public health impact. Population genetics papers were selected on the basis of the compatibility of study design, specifically the analysis of comparable polymorphic systems.

Complementary data on history, linguistics, cultural anthropology and demography were found through standard library and bibliographic searches, and included publications recommended by consulting experts in Romani studies (Drs. Elena Marushiakova and Vesselin Popov from the Bulgarian Academy of Sciences and Dr. Ian Hancock from the University of Texas at Austin).

Genetic studies of the Roma have been conducted for over 70 years, with thousands of individuals sampled across Europe. During the years of the Third Reich, Gypsies, together with Jews, attracted the special attention of German geneticists [9]. A grant proposal signed by Nobel prize winner Ferdinand Sauerbruch and funded by the Deutsche Forschungsgemeinschaft designed the "genetic and medical research" at the death camp in Auschhwitz [9]. The Race Hygiene and Population Biology Research Centre, established in 1936, organised thorough records of Jewish and Romani pedigrees and provided "the scientific basis" for the "final solution", the annihilation of millions of Jews and Roma in the concentration camps of Nazi-occupied Europe.

Post-war genetic research has been preoccupied with the Indian origins of the Roma [[10–16]], pursuing the "Indian connection" even in studies meant to focus on severe genetic disorders [17]. Most studies have remained in the realm of scientific exploration, away from the health needs of the Roma. Many publications display judgemental and paternalistic attitudes, that would be considered unacceptable if used with regard to other populations.

This historical "track record", the persisting practices of discrimination and marginalisation [[3–6]], and the fact that, unlike the Jews, the Finns and the French Canadians, the Roma are still the "object" of investigations conducted by outsiders, are all likely to impact on the attitudes of the Roma towards genetics. Building up the trust and collaboration necessary for both public health programs and research, should become a goal of the health care systems of Europe.

Population genetic studies have used mostly "classical" polymorphisms to investigate Romani individuals from different European counties and address three main questions: (i) similarity between Roma and Indians; (ii) relatedness to European populations; (iii) affinities between Romani populations from different countries [[10–24]]. Single locus comparisons have resulted in controversy, with some pointing to close genetic affinity between Roma and Indians, and others indicating that the Roma are indistinguishable from Europeans. Heterogeneity between countries has become apparent and has led to the conclusion that the European Roma are composed of two different populations, characterised respectively by a high and a low frequency of blood group B [23], or defined as East and West European Roma, with the former closely related to Indian populations [16]. Heterogeneity of Romani populations within the same country has been suggested by the very small number of studies addressing this issue [19,21,25,26].

In an attempt to summarise the existing data, we have conducted a multilocus re-analysis of several marker systems using comparable studies of the Roma in different European countries [11,22,23,24,26,27], Europeans [[28–33]] and north Indian populations [[34–36]]. The comparisons (Figure 2, Table 1) provide a general indication that most of the Roma are genetically closer to Indians than to European populations, hardly surprising in view of the linguistic theories on the Indian origins of the Roma proposed two centuries ago [1]. More importantly, the analysis highlights the internal diversity of the Roma, who appear to be genetically far more heterogeneous than autochthonous European populations.

The pattern emerging from genetic research is that of a conglomerate of founder populations which extend across Europe but at the same time differ within individual countries, and whose demographic history, internal structure and relationships are poorly understood. An insight is provided by the social sciences.

The 18^th century theory on the Indian origins of the Roma [reviewed in 1], is based on the similarities between Romani and languages spoken in the Indian subcontinent and is supported by genetic evidence. However the lack of close relationship to any specific language or dialect has left unresolved the question of the original ethnic composition of the proto-Roma, with both single [72,73] and diverse [74] origins proposed by linguists. Translated into the language of genetics, this is a relevant question related to the homogeneity or diversity of the founding population.

Inferred from linguistic influences retained in all Romani dialects, the major migration routes pass through Persia, Armenia, Greece and the Slavic-speaking parts of the Balkans [1]. The first documents pointing to the arrival of the Roma in the Balkans date from the 11^th-12^th century [1,75]. By the 15^th century, mention of their presence can be found in historical records from all parts of Europe [1,2].

During its subsequent history in Europe, this founder population split into numerous socially divided and geographically dispersed endogamous groups, with historical records from different parts of the continent consistently describing the travelling Gypsies as "a group of 30 to 100 people led by an elder" [1,2]. These splits, a possible compound product of the ancestral tradition of the jatis of India, and the new social pressures in Europe (e.g. Gypsy slavery in Romania [76] and repressive legislation banning Gypsies from most western European countries [1,2]), can be regarded as secondary bottlenecks, reducing further the number of unrelated founders in each group. The historical formation of the present-day 8 million Romani population of Europe is therefore the product of the complex initial migrations of numerous small groups, superimposed on which are two large waves of recent migrations from the Balkans into Western Europe, in the 19^th - early 20^th century, after the abolition of slavery in Rumania [1,2,76] and over the last decade, after the political changes in Eastern Europe [7,8].

The Group is still the primary building block of the social organisation of the Roma [1,2]. Group identity and the ensuing divisions and rules of endogamy are based on tradition, customs and organs of self-rule, language and dialects, trades, history of migrations, and religion. Individual groups can be classified into major metagroups [1,2,75]: the Roma of East European extraction; the Sinti in Germany and Manouches in France and Catalonia; the Kaló in Spain, Ciganos in Portugal and Gitans of southern France; and the Romanichals of Britain [1]. The greatest diversity is found in the Balkans, where numerous groups with well defined social boundaries exist. The 700-800,000 Roma in Bulgaria belong to three metagroups, comprising a large number of smaller groups [75].

Linguistics, history and cultural anthropology suggest two major, equally plausible historical scenarios that could lead to a "jigsaw puzzle" of founder populations: (i) a genetically substructured ancestral population, where the old social traditions of strict endogamy have been retained and subsequent splits of the comprising groups have enhanced the original genetic differences; (ii) a small homogeneous ancestral population spawning numerous subgroups where strong drift effects have resulted in substantial genetic divergence. Genetic research has indeed faced the "jigsaw puzzle" and has thus far been unable to resolve it. The genetic data provide evidence of population stratification, however a closer examination is precluded by the random cross-section sampling design of most population genetic studies, where the traditional social organisation and self-identity of the Roma have been ignored and subjects classified on the basis of the political boundaries of Europe. The relationship between social organisation and genetic structure does not appear to be straightforward and is still to be addressed in population genetic research based on the long standing identity of Gypsy groups. The issue is of relevance to public health policies and the targeted prevention of mendelian disorders, as well as to future studies of genetically complex disorders.

The existing information on single gene disorders is certainly not exclusive to the Roma. The phenomenon of clustering of rare disorders and private founder mutations has been studied in detail in well characterised founder populations, such as the Jews [77,78], Finns [79,80] and French Canadians [81]. Unlike the above examples however, genetic studies of the Roma have failed to take the immediate benefits of research back to the individuals and families that have been the object of research. Yet by now it should be obvious that genetics has an important role to play in improving the quality of health care for the Roma. Treatable disorders such as galactokinase and MCAD deficiency, with an expected incidence of affected births in the range of 1:1,000 to 1:5,000, meet the standard criteria for newborn screening more than does phenylketonuria, with its average incidence of 1:10,000. Adding the simple, sensitive and specific mutation tests to existing newborn screening programs would be technically simple and highly efficient due to the homogeneous genetic basis of the disorders.

Carrier testing should be made available to Romani communities at high risk for severe untreatable disorders. Information on the identity of affected Romani populations is important for public health intervention since it would allow the planning and facilitate the implementation of targeted prevention programs, especially in the Eastern European countries where economic resources are limited. The importance of the educational component of such programs has already been demonstrated by the highly successful prevention of Tay-Sachs disease among Ashkenazi Jews [82] and the failure of sickle-cell screening among Afro-Americans [83]. This component would be particularly important for a population like the Roma, which has been subject to racism and persecution throughout its co-existence with European societies.

The attention of geneticists is increasingly attracted by genetically isolated populations in the third world. In terms of living standards and the major health indicators, the Roma are much closer to the developing world than to their European neighbours [3]. This forgotten people of Europe can be regarded as a test case for the capacity of genetics to provide better health.