Article Text
Abstract
Background Pain, physical function (PF) and quality of life (QoL) are important aspects of the lived experience of those with chronic disease1.
Objectives A EULAR funded study to adapt the Osteoarthritis Quality of Life Scale (OAQoL) into five languages (German, Hungarian, Italian, Spanish, Turkish) gave the opportunity to explore the relationships between these key attributes in osteoarthritis (OA).
Methods A Path Model was employed, within a Structural Equation Framework, utilising Rasch-transformed latent estimates as single indicator latent variables. Measures included lower & upper limb joint count (Manikin), pain on function (WOMAC), PF (HAQ) and QoL (OAQoL). The hypothesis was that the impact of pain upon QoL would be partially mediated by PF and this relationship will be invariant by age, gender and country.
Results 677 patients with OA were recruited across 5 countries, approximately 20% from each. 76% were female, 33% aged 60 and under. 65% were married, 22% in work. Duration differed significantly across countries, with 10.1% of those in Turkey with duration of 15 years or more, compared with 32.5% of those in Spain. Joint count also varied significantly, e.g. 33.8% of those recruited in Spain reported no lower limb involvement compared with just 1.5% in Hungary. Path model fit was acceptable (χ2 7.52 (df 4) p=0.111; RMSEA 0.036; TLI 0.992, CFI 0.997) (Figure). All path weights were consistent with the basic hypothesis and 43% of the impact of pain upon QoL was mediated through PF (Sobel-Goodman Z 10.93; p<0.001). The number of upper & lower joints influenced PF and pain respectively. Thus an increase in pain, and greater limitation in PF gave a worse QoL. The total effect of pain (direct & indirect) was the strongest influence upon QoL. The Wald Test for group invariance showed no significant difference of regression weights (p>0.01) by age, gender, duration and, to a large extent, by country. In the latter there were slight variations in the magnitude of paths weights between pain and PF. The effect appeared much stronger in Italy and Turkey than elsewhere. Overall, the model appeared robust across countries (Model fit χ2 26.85 (df 20) p=0.14; RMSEA 0.05; TLI 0.985, CFI 0.993). Country-specific adjusted R2 for explained variance in QoL ranged between 0.59–0.71.
Conclusions Pain, PF and the consequent reduction in perceived QoL are at the centre of the lived experience of those with OA. The relationship between these three variables, including the mediating role of PF, was found to be consistent across age, gender and, with slight variations, country. This “standard model” for OA opens up the opportunity to further explore potential psychological mediating factors which may improve QoL as well as moderating factors such as different types of intervention.
Parker L, et al. The burden of common chronic disease on health-related quality of life in an elderly community-dwelling population in the UK. Fam Pract 2014;31:557–63.
Disclosure of Interest None declared