(ALI/ARDS) usually develops in response to a major insult such as sepsis, trauma, pneumonia, and multiple transfusions. Our current understanding of why some patients develop and die from ALI while others do not is incomplete. Although genetic determinants to the development of and outcome in ALI/ARDS have only recently been investigated, there has been a sudden explosion of studies in this field in the last few years.

In their review, Gao and Barnes (4) make their latest contribution by summarizing the recent advances in the genetic susceptibility to ALI/ARDS. Their review highlights how genetic association studies in ALI/ARDS can add to our current understanding of this devastating condition. Yet it also makes it clear that the investigation into the genetic susceptibility to ALI/ARDS is still in its infancy and much still needs to be done. One vital need in this field is more replication studies in the future. A causal genotype-phenotype relationship cannot be established with one initial report. As such, replication studies are the backbone to the genetic epidemiology of complex diseases (3). In one meta-analysis, 20% of these replication studies showed a significant result confirming only 44% of these genetic associations (10). But there is debate as to what constitutes a successful replication or refutation of a genetic association in complex diseases. Interpretations of replication studies are challenged by differences in genotyping techniques, insufficient power, and heterogeneity in the study population. Some early studies in ALI/ARDS, like many other early genetic epidemiology studies of complex diseases, focused on reportedly functional variants of, most commonly, single nucleotide polymorphisms (SNPs). Others have examined haplotypes or haplotype-tagging SNPs, which more comprehensively describes the variation in the gene. However, the ability of replicate studies to confirm the original association depends on whether the genotyping method provides adequate coverage of the genetic variation and the linkage disequilibrium (LD) between the true causal locus and variants genotyped in different studies. In addition, multiple susceptibility loci may exist on the same genetic region (13). Thus, the unit of replication should be the gene, and association with any variant or haplotype on the gene should be considered confirmatory (12).