Table 4.

Analysis of covariance between thin-side CTR, thick-side CTR, Logit CTRR, and associated parameters.

		Analysis 1			Analysis 2			Analysis 3
		F	df	p-value	F	df	p-value	F	df	p-value
Thin-side CTR	Age	63.54	1, 125	<0.001 **	72.18	1, 138	<0.001 **	72.78	1, 139	<0.001 **
	Sex (Female)	0.99	1, 125	0.32	not included			not included
	Dominant hand	0.00	1, 125	0.95	not included			not included
	Thin-side WJD	334.40	1, 8.041	<0.001 **	33.60	1, 1.060	0.10	8.00	1, 136	0.005 **
	Thick-side WJD	114.22	1, 8.041	<0.001 **	46.58	1, 1.030	0.09	8.69	1, 139	0.004 **
	Interaction †	0.01	1, 125	0.92	0.19	1, 138	0.67	not included
		Analysis 1			Analysis 2			Analysis 3
		F	df	p-value	F	df	p-value	F	df	p-value
Thick-side CTR	Age	72.78	1, 125	<0.001 **	86.36	1, 138	<0.001 **	86.22	1, 139	<0.001 **
	Sex (Female)	0.91	1, 125	0.34	not included			not included
	Dominant hand	0.00	1, 125	0.96	not included			not included
	Thin-side WJD	1.83	1, 1.028	0.40	0.72	1, 1.014	0.55	0.24	1, 139	0.63
	Thick-side WJD	6.64	1, 1.023	0.23	23.74	1, 1.007	0.13	21.10	1, 139	<0.001 **
	Interaction †	1.28	1, 125	0.26	0.77	1, 138	0.38	not included
		Analysis 2
		F	df	p-value
Logit CTRR	Age	0.41	1, 138	0.52
	Thin-side WJD	2.95	1, 1.003	0.34
	Thick-side WJD	0.30	1, 1.001	0.68
	Interaction †	3.94	1, 138	<0.05 *

CTR: cortical thickness rate. In each patient, the side with the lower cortical thickness rate (CTR) of the metacarpals was designated the “thin-side” and that with the higher CTR the “thick-side”. WJD: wrist joint damage. CTRR: cortical thickness rate ratio = CTR (thin-side)/CTR (thick-side). Logit CTRR = ln {CTRR/(1-CTRR)}. To approximate a normal distribution for analysis of covariance (ANCOVA), CTRR was transformed using logit transformation. This involved applying the logit function to CTRR values, defined as: Logit CTRR = ln {CTRR/(1-CTRR)}. Interaction ^†: interaction between thin- and thick-side WJDs. Analysis 1: ANCOVA with the following dependent variable and factors: age, sex (female), dominant hand, WJD thin-side, WJD thick-side, and interaction between thin- and thick-side WJDs. Analysis 2: ANCOVA with the following dependent variable and factors: age, WJD thin-side, WJD thick-side, and interaction between thin- and thick-side WJDs. (excluding sex (female) and dominant hand in Analysis 1). Analysis 3: ANCOVA with the following dependent variable and factors: age, WJD thin-side, WJD thick-side (excluding interaction between thin- and thick-side WJDs in Analysis 2). F: F-value. df: degrees of freedom, expressed as F (df1, df2), where df1 is df for the factor between-groups and df2 is df for the error within-groups. *: significant, p-value < 0.05. **: significant, p-value < 0.01.