Table 7.

Multivariate logistic regression analyses with OxAFQ-C Footwear domain scores as the dependent variable ^a

	OxAFQ-C Footwearbdomain score
Independent variables	OR (95% CI)c	Change in NagelkerkeR2	pvalue
Foot structure
Arch Indexd	1.03 (0.89 to 1.18)	0.004	0.712
Hallux valgus (presence)	3.08 (0.23 to 41.79)	0.012	0.399
Lesser toe deformity (presence)	2.72 (0.17 to 43.62)	0.014	0.480
Footwear fite
Length (ill-fitting/lower median)f	0.23 (0.02 to 2.70)	0.061	0.240
Width (ill-fitting/lower median)f	0.89 (0.11 to 7.15)	0.042	0.915

^aAdjusted for sex, age and body mass index.

^bOxAFQ-C Footwear domain scores were dichotomised where the lowest quartile (item responses: always, very often) was classified as ‘disability’ and the upper three quartiles (item responses: never, rarely and sometimes) were classified as ‘no disability’.

^cOR: Odds ratio.

^dArch Index multiplied by hundred.

^eLogistic regression analyses performed for footwear fit were based on n = 48 due to missing data.

^fIll-fitting: representing the greatest disparity between the participant’s foot and footwear length and width dimensions; expressed as a percentage. The values that defined the lower median (ill-fitting) were as follows: length (shoe 14.3% shorter to 10.8% longer than foot), width (shoe 26.5% to 4.9% narrower than foot).