1. ¹Oxford Centre for Diabetes, Endocrinology and Metabolism, Churchill Hospital, Oxford, UK
2. ²Department of Epidemiology, Imperial College (St. Mary's Campus), London, UK
3. ³Department of Epidemiology and Public Health, Imperial College (St. Mary's Campus), London, UK
4. ⁴Department of Clinical Chemistry, Oulu University Hospital and University of Oulu, Oulu, Finland
5. ⁵Department of Obstetrics and Gynecology, Oulu University Hospital and University of Oulu, Oulu, Finland
6. ⁶Department of Child and Adolescent Health, National Public Health Institute, Finland
7. ⁷Department of Public Health Science and General Practice, Oulu University Hospital and University of Oulu, Oulu, Finland
8. ⁸Department of Diabetes Research & Vascular Medicine, Peninsula Medical School, Exeter, UK
9. ⁹Wellcome Trust Centre for Human Genetics, Oxford, UK
10. ¹⁰Institute of Reproductive and Developmental Biology, Imperial College (Hammersmith Campus), London, UK

Correspondence: Dr TM Barber, Diabetes Research Laboratories, Oxford Centre for Diabetes, Endocrinology and Metabolism, Churchill Hospital, Old Road, Headington, Oxford OX3 7LJ, UK. Tel: +44 1865 857292; Fax: +44 1865 857299; E-mail: tom.barber@drl.ox.ac.uk

Received 6 November 2006; Revised 24 January 2007; Accepted 1 February 2007; Published online 7 March 2007.

Abstract

Polycystic ovary syndrome (PCOS) is strongly associated with hyperinsulinaemia and type II diabetes (T2D). Sequence variation within KCNJ11 (encoding Kir6.2, the beta- cell inwardly rectifying potassium channel) is implicated in the pathogenesis of neonatal diabetes, hyperinsulinaemia of infancy and multifactorial T2D. Comprehensive tagging studies have demonstrated that the KCNJ11 E23K variant (or ABCC8 A1369S in LD>0.9) is responsible for the known association between KCNJ11 and T2D. Given the phenotypic overlap between PCOS and T2D, we investigated whether E23K is involved in susceptibility to PCOS and related traits. Case–control analyses for the KCNJ11 E23K variant were performed in (a) 374 PCOS cases and 2574 controls of UK British/Irish origin, and (b) 550 women with PCOS symptoms and 1114 controls from a Finnish birth cohort. The relationship between E23K genotype and androgen levels (a key intermediate phenotype relevant to PCOS) in 1380 samples was studied. The UK case–control analysis revealed no association between E23K genotypes and PCOS status (P=0.49; Cochran–Armitage test), and no significant relationship between E23K genotype and androgen measures in the samples for which these phenotypes were available (P=0.19). Similarly, the Finnish case–control analysis showed no association between E23K genotypes and PCOS status (P=0.75; Cochran–Armitage test), and no significant relationship between E23K genotype and androgen measures in the samples for which these phenotypes were available (Finnish controls, P=0.25; Finnish cases, P=0.08). In conclusion, these data (involving >4600 subjects) provide no evidence that common variants of the KCNJ11 E23K polymorphism have a major influence on PCOS susceptibility, though modest effect sizes (OR<1.25) cannot be excluded.