Mode of inheritance of hand osteoarthritis in ethnically homogeneous pedigrees

Human Biology, Dec 2002 by Livshits, Gregory, Kalichman, Leonid, Cohen, Zvi, Kobyliansky, Eugene

The population of Chuvasha is characterized by demographically stable family structures with long-time "traditional" relations between family members. They have lived under the same environmental conditions for at least the last three generations and have not been exposed to outside influences, such as genetic flow (Tischkov 1994). Most of families share similar living, economic, and occupational conditions. The information that was collected on the family members included sex, age, and occupation as well as nature and extent of physical activity. Data on chronic morbidity and medical treatment were also included in the questionnaire and compiled for this study. Individuals with known bone disease, hormone replacement therapy, steroid medicine intake, amenorrhea, or posttraumatic, rheumatoid, or psoriatic arthritis were excluded from the study. Measurements of standard anthropometrical traits such as stature and body mass were carried out and recorded. Finally, X-ray films of both hands were obtained for each individual.

Roentgenographic Assessment of OA. For studying the genetic predisposition to a complex disease such as OA, accurate phenotype ascertainment is a necessity, and radiological changes are considered superior to clinical manifestations in the diagnostic definition of OA (Hart et al. 1994). Single X-ray radiographs from both hands were taken from each subject. The extent of OA development was evaluated separately for 14 joints of each hand, i.e., five distal interphalangeal (DIP), four proximal interphalangeal (PIP), and five metacarpophalangeal (MP). The Kellgren and Lawrence grading scheme, which utilizes photographs from the Atlas of Standard Radiographs (Kellgren and Lawrence 1963), was used in the present study. The development of osteophytes, joint space narrowing, subchondral sclerosis, lateral deformity, and cortical collapse were taken into account in assigning a grade for each of the 14 joints. The extent of the OA development at each joint ranged between 0 and 4 grades, yielding a possible total range of score between 0 and 56 in each hand.

Statistical and Genetic Analyses. The total individual OA score was used in the statistical genetic analysis. It was obtained by means of the principal component analysis (PCA) of two hands for rows of the DIP, PIP, and MP joints. The first principal component derived from the outcome of this analysis (PC1-OA) was then used in further analyses. Univariate analyses were undertaken to investigate interrelationship between PC1-OA and age, sex, body height, and weight. Statistical analysis was performed using the STATISTICA 5.5 package for Windows (StatSoft 2000). We computed the familial correlations of PC1-OA, adjusted for significant covariates, by means of the "Statistical Analysis for Genetic Epidemiology" (SAGE release 3.1) program.

The SA, as it is implemented in the statistical package MAN (Ginsburg 1997), was undertaken in this study to test a major gene model of inheritance of hand OA. The mixed model of inheritance (Morton and MacLean 1974; Elston 1981; Lalouel 1983; Beaty 1997; Ginsburg and Livshits 1999) had estimated effects of a potential major gene and possible multifactorial effects, and is described in detail in numerous publications, including the aforementioned. The parameters that had been estimated in the general model are defined as follows: P^sub 1^ and P^sub 2^ are the population frequency of homozygous and heterozygous genotypes (A^sub 1^A^sub 1^ and A^sub 1^A^sub 2^); [mu]^sub gs^ is the average genotypic value in all individuals having genotype g and sex s; g = 1, 2, and 3 corresponds to genotypes A^sub 1^A^sub 1^, A^sub 1^A^sub 2^, and A^sub 2^A^sub 2^, respectively; [sigma]^sup 2^^sub g^ is the trait variance in individuals having the same major gene genotype g; it estimates the trait variation due to the effect of all possible environmental factors and potential minor genes; [tau]^sub g^ is the transmission probability parameter (it estimates the probability that a parent of genotype g transmits allele A^sub 1^ to the next generation); [rho], [beta], and [epsilon] are partial correlations between the trait residuals adjusted for a major gene effect, in spouses, parents/offspring, and the siblings, respectively; and, finally, T^sub gs^ and T^sup 2^^sub gs^ are linear and quadratic genotype/sex specific coefficients of age dependence. A maximum likelihood ratio test was used as a model-fitting technique for the comparison between the general model and a more limited model, i.e., one containing one or more parameters constrained to the expected value (the number of degrees of freedom [df] represents the number of constrained parameters). A best-fitting and most parsimonious model was established after dropping all nonsignificant parameters from the general model.