Multisite joint pain in older Australian women is associated with poorer psychosocial health and greater medication use

Table 3 Generalised ordinal logistic regression analysis of the variables associated with none, some and many joint pain sites across pain, health, and sociodemographic variables

	No joint pain compared to multisite joint pain (some and many joint pain sites)			No joint pain and some joint pain compared to many joint pain sites
	Odds Ratio	95% CI	p value	Odds Ratio	95% CI	p value
Number of prescribed medications	1.08	1.00, 1.16	0.05	1.08	1.00, 1.16	0.05
Duration of joint pain	1.00	1.00, 1.00	< 0.01	1.00	1.00, 1.00	< 0.01
Present Pain Intensity
Discomfort	0.74	0.3, 1.84	0.52	2.35	1.41, 3.93	< 0.001
Distressing/Horrible/Excruciating	5.09	1.75, 14.78	< 0.01	5.09	1.75, 14.78	< 0.01
SF-36 physical component scale	1.97	0.98, 1.01	0.15	0.95	0.92, 0.98	< 0.001

Abbreviations: CI confidence interval, SF-36 PCS Medical Outcomes Study 36 Item Short Form Survey Physical component scale
In the final reduced model of generalised ordinal logistic regression, coefficients for the change in odds of having multisite joint pain (comparing no joint pain to the combination of some and many joint pain sites (column two); and being in a higher joint pain category (comparing the combination of no joint pain and some joint pain sites to many joint pain sites (column three)) for each predictor and adjusted for other predictors in the model are shown

ISSN: 2045-709X