Osteoarthritis and Frailty in Elderly Individuals Across Six European Countries

Results From the European Project on OSteoArthritis (EPOSA)

Maria Victoria Castell; Suzan van der Pas; Angel Otero; Paola Siviero; Elaine Dennison; Michael Denkinger; Nancy Pedersen; Mercedes Sanchez-Martinez; Rocio Queipo; Natasja van Schoor; Sabina Zambon; Mark Edwards; Richard Peter; Laura Schaap; Dorly Deeg


BMC Musculoskelet Disord. 2015;16(359) 

In This Article


This study suggests a strong association between OA and frailty that remains after adjusting for socio-demographic and health-related variables and that is maintained when analysing different sites separately (knee, hip, hand). The odds of pre-frailty and frailty were 1.54 and 2.96 times higher among OA than non-OA patients.

Recent publications have found an independent association between hip OA and frailty or pre-frailty in men aged 65 and over[27] and knee OA has been shown to be associated with a greater prevalence and risk of developing frailty.[20] Other studies have found a relationship between frailty and OA measured with subjective criteria.[15,21] These results suggest common pathophysiological mechanisms underlying both conditions. Certain inflammatory cytokines (IL1, IL-6 and TNFalpha) that are involved in the frailty cycle[18,28,29] are increased in OA cartilage as opposed to normal cartilage. The response to growth factors such as IGF1 declines markedly, thus inhibiting maintenance of normal cartilage and promoting the development of OA.[18,28]

The high prevalence of OA among the elderly is well known.[3–7] In our study, 30.4 % of cases had OA in one or more of the joints studied. The prevalence of OA at any site was very high in Italy (42.3 %) and was low in Germany (19.7 %), and these differences were maintained regardless of the joint affected. Although there are demographic differences across countries such as age, sex, educational level, the variability in prevalence rates may be influenced by other factors as climate, health care, lifestyle or environmental factors.[22,30] The findings of other studies in the developed world are consistent with ours. In the Rotterdam Study the prevalence of OA based on clinical criteria was 16.2 % in men and 20.4 % in women aged 75–84 years.[31] The Johnston County study reported a frequency of 47 % in men and 49.2 % in women aged 70–74 years.[32]

The prevalence of frailty is high in the elderly and will increase in the future due to the progressive ageing of the population. Our results show that the overall prevalence of frailty is 10.2 % and of pre-frailty is 51.0 %, which is consistent with other population-based studies of similar characteristics conducted in Europe.[14,21] The Survey of Health, Ageing and Retirement in Europe (SHARE),[20] which compared frailty across 10 European countries, found a much higher prevalence in the Mediterranean countries (Italy and Spain) and a lower prevalence in Nordic countries like Sweden. Because of the high frequency and the earlier stage of the frailty spectrum, the prefrail constitute a target population requiring action from the health and social services.[33,34]

Obesity is strongly associated with OA and frailty in older people, and in the presence of the three conditions there is a higher risk of functional limitation.[19,27,35–37] In our study, obese individuals had a higher risk of being prefrail and frail, as we can see in Table 3. It is important to realize that obesity acts not only as a local biomechanical factor, but as a systemic component[4] and its influence increases with age: Body composition changes with age, even if body mass index (BMI) does not vary, with an increased proportion of fat mass and decrease in lean mass. It is now appreciated that these age-related changes occurring in tissues besides articular cartilage may contribute to the development of OA.[1] However, these changes are much more intense in the presence of obesity, frailty and/or OA and, as noted by several authors, may be due to the development of insulin resistance and the maintenance of chronic inflammatory processes over time.[8,10,14] Sarcopenic obesity, a condition in which lean body mass is lost while fat mass may be preserved or even increased[17] has a stronger association with knee OA than non-sarcopenic obesity, indicating the importance of the systemic metabolic effect of obesity in OA.[29]

This study has some limitations. Although all participants were recruited from pre-existing community-based cohorts of older individuals, the cohorts may not be representative of their respective countries to the same extend, due to differential attrition. However, these differences may allow hypothesis generation regarding the association between OA and frailty. The non-participation in the 1-year follow-up of 487 baseline individuals may influence the results, but given the higher proportion of women and older individuals, which are variables associated with a higher prevalence of frailty and OA, the results presented would likely have been stronger had we been able to include the non-participants. Another limitation is that the involvement of multiple research centres meant that data collection methods might vary by study site. However, all questionnaires and protocols for examination were undertaken by one team to minimize such problems and OA was diagnosed according to clinical criteria and following a similar methodology in all the participating cohorts. Finally, although the study design is longitudinal, given that the main variables (OA and frailty) are chronic processes, and that the period between baseline and the follow-up wave was very short, we cannot assume the temporal direction of causality in the association detected. Other longitudinal studies over a longer time period are needed to assess the relationship between the incidence of frailty or pre-frailty and the presence of OA.

Due to the high prevalence of OA and frailty in persons aged 65 and over, the strong association between the two processes and the fact that frailty has been considered as a predictor of mid-term mortality in individuals with OA,[35] several authors have recommended preventive and therapeutic interventions at the community level. These include, firstly, the early detection of frailty as an important measure of overall health status in the elderly.[27] The frailty score proposed by Fried et al. has biological validity and is easy and inexpensive to measure.[14] Furthermore, OA seems to be associated with worsening of and/or lower recovery from frailty.[33] Thus, we should encourage the use of this scale in the screening of older persons, especially those age 75 and over.[38]

Another intervention may be to promote or prescribe appropriate exercise. The practice of regular physical activity that includes both aerobic and resistance exercises tailored to each individual's needs is probably the most efficient intervention to prevent frailty and to delay disability and the adverse events associated with obesity and frailty.[34,39,40] Health benefits can be obtained at any point in the evolution of OA and frailty.[40]

Finally, weight loss is a priority in the long-term management of obese individuals with OA. Not only does this help lessen joint overload, it also results in decreased fat mass and a proportional increase in lean mass which leads to functional improvement.[29] The Mediterranean diet, rich in olive oil, fruits and vegetables, and with a predominance of polyunsaturated fatty acids, also contributes to improved functionality and quality of life in individuals with osteoarticular pathology.[41]