Table 4:

Top factors associated with the Myositis Damage Index score at cross-sectional evaluation in 49 adult patients with juvenile idiopathic inflammatory myopathies.

	Bivariate Analysis		Multivariable Regression
Predictors of MDI Score	Corr. Coef.^a	P value	Est^b	SE	P value^c
Erythroderma (ever)	0.42	0.003	3.94	1.50	0.01
Shawl sign (ever)	0.35	0.01	2.71	0.94	0.006
Wheelchair (ever)	0.31	0.03
Disease duration (per year)	0.31	0.03	0.08	0.04	0.04
Age at diagnosis (per year)	−0.30	0.04
Falling episodes (ever)^d	0.34	0.04
Gottron’s papules (ever)	0.29	0.04
Clinical subgroup (JDM, JPM, JCTM)	NA	0.04
Worst ACR functional class (per level)	0.28	0.05	1.31	0.37	0.001
Azathioprine use (ever)	0.28	0.05
Disease course	NA	0.05
Ulceration (ever)	0.25	0.09
Malar rash (ever)	0.23	0.11
Antinuclear antibodies’^d	0.26	0.12
Cuticular overgrowth (ever)	0.22	0.13
Heliotrope rash (ever)	0.22	0.13	2.06	1.01	0.05
Number of DMARDs used	0.21	0.14

Point-biserial correlation coefficient for dichotomous variables; Spearman correlation coefficient for ordinal or continuous variables; correlation coefficient not applicable for 3-level categorical variables (p-value based on analysis of variance test).

Estimated slope (b)

P-values are not adjusted for model selection and therefore may overstate the significance of the associations.

For falling episodes, N = 37; for antinuclear antibodies, N = 38. Due to missing data, these variables were not included in the multivariable model.

Blank cells represent variables that were removed during backward selection of the multivariable model.

Abbreviations: ACR: American College of Rheumatology; DMARDs: Disease-Modifying Anti-Rheumatic Drugs; JCTM: juvenile connective tissue overlap myositis; JDM: juvenile dermatomyositis; JPM: juvenile polymyositis; NA: not applicable; MDI: Myositis Damage Index; SE: standard error