Objective: To establish a mathematical model to predict the probability of symmetry of joint involvement as a function of the number of joints involved and to compare expected with actual probabilities in psoriatic arthritis (PsA) and rheumatoid arthritis (RA) and in early and late disease.
Methods: Random involvement of joints was assumed, and the binomial theorem was used to give the frequency distribution of involved joints as a function of each joint count. Ten joint pairs were included: shoulder, elbow, wrist, metacarpophalangeal joints, proximal interphalangeal (PIP) joints of the hands, hip, knee, ankle, metatarsophalangeal joints, and PIP joints of the feet. Observed probabilities were obtained from subjects with early (duration < or =12 months) and late PsA and RA.
Results: The number of subjects in each of the disease subgroups was as follows: early PsA n = 33, late PsA n = 77, early RA n = 61, late RA n = 93. Observed probabilities of symmetry exceeded predicted probabilities for all disease subgroups. The median number of involved joints in each group was as follows: early PsA 4, late PsA 8, early RA 8, late RA 15 (chi2 = 95.3, 3 degrees of freedom, P = 0.0001, by Kruskal-Wallis test). After correcting for the discrepancy in the number of involved joints, no difference in joint symmetry was found between the groups (chi2 = 1.77, P = 0.62 by Friedman two-way analysis of variance). Similar results were obtained when individual hand and foot joints were analyzed separately.
Conclusion: The pattern of joint involvement is often used to distinguish between rheumatoid and psoriatic arthritis. This study confirms that symmetry is largely a function of the total number of joints involved and that, in terms of joint pattern, differences between these disorders are more quantitative than qualitative. Both disorders have high absolute values of symmetry, particularly in the joints of the wrist and hand.