Background: Although there have been suggestions that serum paraoxonase is important in protecting against coronary heart disease (CHD), a large number of studies of genetic determinants of serum paraoxonase have reported apparently conflicting results about their association with CHD.
Methods: We conducted a meta-analysis of 43 studies of the Q192R, L55M, and T(-107)C polymorphisms in the paraoxonase PON1 gene and the S311C polymorphism in the PON2 gene (all of which are in moderately strong linkage disequilibrium with one another), involving a total of 11212 CHD cases and 12786 controls. We explored potential sources of heterogeneity.
Findings: In a combined analysis of all studies, the per-allele relative risk of R192 for CHD was 1.12 (95% CI 1.07-1.16), but in the five largest studies it was only 1.05 (0.98-1.13). Combined analyses of studies of the M55, (-107)T, and C311 variants showed no significant overall associations with CHD, yielding per-allele relative risks of 1.00 (0.95-1.06), 1.02 (0.92-1.14), and 1.04 (0.93-1.17), respectively.
Interpretation: In contrast to previous suggestions, this meta-analysis shows no significant association of CHD with the L55M or T(-107)C polymorphism in PON1 or with the S311C polymorphism in PON2. The weak overall association between the Q192R polymorphism and CHD is of uncertain relevance, particularly since there was no significant association among the larger studies which should be less prone to selective publication. These findings reinforce the need for much larger and more rigorous investigations of the genetic determinants of complex diseases than is now customary, as well as for regularly updated systematic appraisals of such studies to help improve interpretation and prioritise hypotheses.