Iron homeostasis dysregulation has been regarded as an important mechanism in neurodegenerative diseases. The H63D and C282Y polymorphisms in the HFE gene may be involved in the development of sporadic amyotrophic lateral sclerosis (ALS) through the disruption of iron homeostasis. However, studies investigating the relationship between ALS and these two polymorphisms have yielded contradictory outcomes. We performed a meta-analysis to assess the roles of the H63D and C282Y polymorphisms of HFE in ALS susceptibility. PubMed, MEDLINE, EMBASE, and Cochrane Library databases were systematically searched to identify relevant studies. Strict selection criteria and exclusion criteria were applied. Odds ratios (ORs) with 95% confidence intervals (CIs) were used to assess the strength of associations. A fixed- or random-effect model was selected, depending on the results of the heterogeneity test. Fourteen studies were included in the meta-analysis (six studies with 1692 cases and 8359 controls for C282Y; 14 studies with 5849 cases and 13,710 controls for H63D). For the C282Y polymorphism, significant associations were observed in the allele model (Y vs C: OR=0.76, 95%CI=0.62-0.92, P=0.005) and the dominant model (YY+CY vs CC: OR=0.75, 95%CI=0.61-0.92, P=0.006). No associations were found for any genetic model for the H63D polymorphism. The C282Y polymorphism in HFE could be a potential protective factor for ALS in Caucasians. However, the H63D polymorphism does not appear to be associated with ALS.

ABSTRACT. The accuracy of quantitative trait loci (QTLs) identified using different sample sizes and marker densities was evaluated in different genetic models. Model I assumed one additive QTL; Model II assumed three additive QTLs plus one pair of epistatic QTLs; and Model III assumed two additive QTLs with opposite genetic effects plus two pairs of epistatic QTLs. Recombinant inbred lines (RILs) (50-1500 samples) were simulated according to the Models to study the influence of different sample sizes under different genetic models on QTL mapping accuracy. RILs with 10-100 target chromosome markers were simulated according to Models I and II to evaluate the influence of marker density on QTL mapping accuracy. Different marker densities did not significantly influence accurate estimation of genetic effects with simple additive models, but influenced QTL mapping accuracy in the additive and epistatic models. The optimum marker density was approximately 20 markers when the recombination fraction between two adjacent markers was 0.056 in the additive and epistatic models. A sample size of 150 was sufficient for detecting simple additive QTLs. Thus, a sample size of approximately 450 is needed to detect QTLs with additive and epistatic models. Sample size must be approximately 750 to detect QTLs with additive, epistatic, and combined effects between QTLs. The sample size should be increased to >750 if the genetic models of the data set become more complicated than Model III. Our results provide a theoretical basis for marker-assisted selection breeding and molecular design breeding.