Clinical review: the genetics of type 2 diabetes: a realistic appraisal in 2008

Abstract

Context: Over the last few months, genome-wide association studies have contributed significantly to our understanding of the genetic architecture of type 2 diabetes. If and how this information will impact clinical practice is not yet clear.

Evidence acquisition: Primary papers reporting genome-wide association studies in type 2 diabetes or establishing a reproducible association for specific candidate genes were compiled. Further information was obtained from background articles, authoritative reviews, and relevant meeting conferences and abstracts.

Evidence synthesis: As many as 17 genetic loci have been convincingly associated with type 2 diabetes; 14 of these were not previously known, and most of them were unsuspected. The associated polymorphisms are common in populations of European descent but have modest effects on risk. These loci highlight new areas for biological exploration and allow the initiation of experiments designed to develop prediction models and test possible pharmacogenetic and other applications.

Conclusions: Although substantial progress in our knowledge of the genetic basis of type 2 diabetes is taking place, these new discoveries represent but a small proportion of the genetic variation underlying the susceptibility to this disorder. Major work is still required to identify the causal variants, test their role in disease prediction and ascertain their therapeutic implications.

Figure 1

Chronological accounting of the discovery of type 2 diabetes-associated genes, plotted by year of definitive publication and approximate effect size. Within each group, darker colors indicate biological candidate genes (PPARG, KCNJ11, HNF1B, WFS1, HNF4A, HNF1A, and ENPP1), and lighter colors denote loci identified via agnostic genome-wide approaches. In gray are genes implicated in type 2 diabetes by considerable functional and genetic evidence but for which genome-wide statistical evidence of association has not been reached. Allelic effect sizes are derived from Ref. , except for PPARG (110), KCNJ11 (16), TCF7L2 (67), the six genes identified in 2008 (103), HNF4A (30), HNF1A (31), CAPN10 (51), and ENPP1 (45).

Figure 2

Power estimates of the DIAGRAM metaanalysis (effective n = 9562) to detect genetic associations of different OR (range 1.1–1.4) across various risk allele frequencies. Although power is excellent for effect sizes of about 1.4 (indicating it is unlikely that SNPs of such effects remain undiscovered, at least in populations of European ancestry), power is much more limited for lower effect sizes (suggesting that only a small fraction of comparable associations has been identified).