Strong association between polymorphisms in ANKH locus and skeletal size traits

Abstract

Loss of bone strength is the main determinant of bone fragility. Bone strength is directly dependent on bone size (BS). A substantial portion of BS variation is attributable to genetic effects. However, the list of genes and allelic variants involved in the determination of BS variation is far from being complete. Polymorphisms in the ANKH gene have been shown to be associated with radiographic hand BS-related phenotypes. The present study examined the possible association of the ANKH gene with skeletal size and shape in order to test the universality of the ANKH effect on BS traits. Our sample consisted of a total of 212 ethnically homogeneous nuclear families (743 individuals) of European origin. Each individual was measured for body height, weight, and several other anthropometrical measurements, and genotyped for nine polymorphic markers (the average heterozygosity level was 0.4). We observed significant associations with practically all the anthropometrical phenotypes studied. More specifically, we found associations with body weight and height, limb length (P≤0.001; promoter region). After adjustment for body height, we demonstrated the substantial association increase for biacromial breadth (P=0.0012; some 140 kb downstream from ANKH) and vertebral column length (P=0.0008; exons 2–7 region). The majority of the observed associations persisted even after correction for multiple testing. For the first time the reliable evidence in support of universality of ANKH gene polymorphisms effect on bone size was provided.

Appendix

Multiple comparison procedure

Each point of the simulated distribution was characterized by a vector of one-dimensional P-values (P ₁, P ₂, P ₃, P ₄) with components corresponding to separate TDTs. The computation of the MCP P-value included two steps. First, to compare any pair of distribution points as more or less significant, we determined a function S (P ₁, P ₂, P ₃, P ₄) as a point significance level and computed this level for each distribution point. Second, to estimate the appropriate MCP P-value, P _MCP, for each vector (P ₁₀, P ₂₀, P ₃₀, P ₄₀), we computed the fraction f of distribution points, which were more or equally significant in comparison with the level S ₀=S (P ₁₀, P ₂₀, P ₃₀, P ₄₀). Thus, f was used as an estimator for P _MCP (the probability to reject erroneously the null hypothesis using the S ₀ level as a criterion for rejection).

There are a number of ways to define the significance level function S. For example, we can define it as the minimum of one-dimensional P-values, S (P ₁, P ₂, P ₃, P ₄) = min (P ₁, P ₂, P ₃, P ₄) (simulative analog of Bonferroni correction), or as their maximum. To account for the relationship between all P-values, we used the population level for asymmetrical (A) and symmetrical (B) rejection regions as follows. Define that the distribution point I, having coordinates P _1i , P _2i , P _3i , P _4i , belong to the rejection region of type A, if it satisfies the condition (P _1i <P ₁) & (P _2i <P ₂) & (P _3i <P ₃) & (P _4i <P ₄), where & denotes the logical AND. Let P ⁽¹⁾<P ⁽²⁾<P ⁽³⁾<P ⁽⁴⁾ be the set of arranged P-values regardless of the serial numbers of tests for which each P-value was computed. Define that the distribution point i belongs to the rejection region of type B, if it satisfies the condition (P ⁽¹⁾ _i <P ⁽¹⁾) & (P ⁽²⁾<P ⁽²⁾ _i ) & (P ⁽³⁾ _i <P ⁽³⁾) & (P ⁽⁴⁾ _i <P ⁽⁴⁾). Define S _A (P ₁, P ₂, P ₃, P ₄) as the fraction of distribution points belonging to rejection asymmetrical region A and S _B (P ₁, P ₂, P ₃, P ₄) as the fraction of distribution points belonging to symmetrical rejection region B. Thus, we have two scores for arranging the distribution. Despite different arrangements, the resulting estimated MCP P-values of both types are generally of the same order. We computed both of them in order to use the more conservative estimator as the general adjusted P-value for each tested pair: the trait–dichotomous marker.