Aqui fica a introdução de um estudo que me parece bastante interessante. Leia tudo no link acima.
'Introduction
It has long been noticed that the incidence of Type 1 Diabetes (T1D) is highly variable from one country to another. For example, the 62.42/100.000 persons/year incidence found in Finland [1] was 780-fold larger than the 0.08/100.000 persons/year incidence in Papua New Guinea [2]; differences in T1D incidence are also observed between countries where the health care systems are comparable. The variability of T1D incidence is even visible within countries; for example, in Italy, T1D incidence varied between 54.4/100.000 persons/year in Sardinia [3] and 4.4/100.000 persons/year in Lombardia [4]. The reason for these differences is not precisely known, but is most unlikely due to classification bias, as the disease cannot go untreated, and the diagnosis is relative easy to perform in children [5]. The country-to-country T1D variability is known to be partly explained by genetic variations. Indeed, HLA (human leukocyte antigen) and 33 other genes are associated with elevated risk of T1D (T1Dbase Version 4.18 updated on 30/9/2014 available at http://www.t1dbase.org) [5–11]. The genetic characteristics of several populations have been found to—at least partially—explain the level of their T1D incidence [12]. For example, the low incidence in Japan, and more generally in southeast Asia, was strongly associated with the absence of highly susceptible haplotypes, such as DRB1*03-DQB1*0201 and DRB1*04-DQB1*0302 found in Caucasian populations [13] or DRB1*030101-DQB1*0201 [14] found in Arab populations (Bahrainis, Lebanese, and Tunisians). Instead, the major susceptible HLA haplotypes in the Japanese and Korean populations were DRB1*0405-DQB1*0401 and DRB1*0901-DQB1*0303 [15].
Another peculiarity of T1D epidemiology is that a dramatic increase of the incidence (on average 3% per year [16]) was observed over the last decades in many countries, in particular European countries with previously low incidences. This increase cannot be explained by genetic factors, since the genetic structure of these countries cannot have varied greatly over such a short period of time. The reasons are more likely to be found in environmental factors (taking here environment broadly, as encompassing physical, chemical, social and life-style factors). However, no single environmental factor, or configuration of factors, that could explain the patterns of differences has ever been identified. More likely, there are complex networks of environmental causes, and of gene-environmental causes that remain to be discovered.
The search of genetic factors of T1D was facilitated during the last 10 years by the GWAS (genome-wide association studies) technology that replaced the gene candidate approaches and instead scanned the entire genome to find SNPs (Single-Nucleotide Polymorphisms) that were significantly associated with T1D [17]. The discovery of an SNP was not the discovery of a “gene”, but was a marker leading to the possible discovery of a gene. Here, we translate this data-driven approach to search for environmental markers related to variations of T1D incidence that might eventually lead to environmental causes, possibly in interaction with genetic factors. One could indeed expect that, in this age of information, plenty of environmental characteristics could be readily available, insofar as local and global organizations collect such data, and provide them free to researchers, with easy interface on the Internet. A limitation commonly advocated is that country statistics are in many cases of too low quality. However, for two related reasons, this argument does not hold, or will not hold for long: a) why collect, maintain and publish such statistics if they cannot be used by researchers? and b) how can one encourage a better quality for these statistics—meaning more resources devoted to them—if they are never used?
Here, we present the attempt we made to use open public data to identify climate and environmental, demographic, economic, and health characteristics correlated with variations in T1D incidence between countries.'