Arthrosis cannot be described as a single disease, but rather as a heterogeneous group of diseases with similar clinical manifestations involving common pathological and radiological changes. The OARSI recently defined arthrosis as a disorder which affects the mobile joints that is characterised by cellular stress and degradation of the extracellular cartilage matrix and commences with micro- and macrolesions which activates poorly adapted repair responses, including innate immunity pro-inflammatory routes. The disease manifests firstly as a molecular alteration (abnormal joint tissue metabolism), followed by anatomical or physiological alterations (characterised by cartilage degradation, bone remodelling, joint inflammation and loss of normal joint function), which may culminate in the emergence of the disease.1

Arthrosis is considered to be the most common rheumatological disease; nevertheless, epidemiological studies sometimes show different results. These differences are largely due to the definition of patients with arthrosis. Due to this it is important when interpreting epidemiological data to know whether the studies in question have defined cases as symptomatic or radiological arthrosis.2

Three cross-sectional studies have been published in Spain which calculate the prevalence of arthrosis. The first of these, which was published in 1982, was based on a study undertaken in the town of Los Cortijos (Ciudad Real). A total of 827 individuals were analysed, 87% of the selected population, by means of their clinical history, physical examination, analysis and radiological study. 20% of the population were over the age of 60 years old. 26.1% suffered arthrosis.3 The second study was undertaken in Asturias and published in 1993. 702 individuals over the age of 18 years were evaluated, 76% of the invited population, selected by randomised, stratified, multistage and proportional sampling. All of the subjects were interviewed by trained doctors, and validated criteria were applied to establish the diagnoses. 23.8% of the sample was diagnosed with arthrosis.4 The third study was EPISER2000, promoted by the Sociedad Española de Reumatología, with subjects over the age of 20 years. It estimated that the prevalence of symptomatic knee arthrosis in Spain stood at 10.2% and that of the hands at 6.2%.5,6 It has to be underlined that EPISER2000 did not analyse the prevalence of axial arthrosis.

Sociodemographic and lifestyle changes over recent years in Spain justify updating the epidemiological data corresponding to arthrosis.7 The sociodemographic changes can be summarised as an increase in the proportion of individuals over the age of 64 years and an increase in the percentage of foreign population, according to Instituto Nacional de Estadística data.8 Regarding lifestyle changes, an increase in obesity and overweight individuals has been observed, together with changes in tobacco consumption.9–11

The main aim of the EPISER2016 study, which was promoted by the Sociedad Española de Reumatología, was to estimate the prevalence of the main rheumatic diseases within the adult population in Spain. These diseases include arthrosis in the locations already included in EPISER2000 (the hip, knee and hand) and this update expands to include the cervical and lumbar spinal column. As a secondary objective its associations with sociodemographic, anthropometric and lifestyle variables have also been examined.

The methodology and sample characteristics of the EPISER2016 study have been described before.7,12 To summarise, it is a cross-sectional study of a population base. Multi-stage stratified and grouped sampling was used. Subjects who live in 78 towns of 17 autonomous communities took part, within the catchment area of 21 hospitals (Fig. 1). Given the size of the sample and the limited magnitude of the differences respecting the general population for the characteristics that were analysed, the sample finally included in EPISER2016 could be considered to be representative of the general adult population of Spain for the estimation of the prevalence of rheumatic diseases. The sample included in the arthrosis prevalence study was composed of 3336 individuals aged 40 years and above.

Fig. 1.

Hospitals and towns which took part in each one of the participating autonomous communities: Hospital General Universitario de Elda: Cañada, Algueña, Sax, Beneixama, Biar, El Pinós, Salinas, Hondón de los Frailes, Hondón de las Nieves (Comunidad Valenciana). Hospital General de Ontinyent: Fontanars del Aforins, Atzeneta D’Albaida, Bocairent, Agullent, Aielo de Malferit, Albaida, Bèlgida, Palomar, Benisoda, Carricola, Bufalí (Comunidad Valenciana). Complejo Asistencial Universitario de Palencia: Torquemada, Baltanás, Paredes de Nava, Dueñas, Villada, Herrera de Pisuerga, Carrión de los Condes, Saldaña, Osorno (Castilla y León). Hospital Comarcal de Monforte de Lemos: Quiroga, Carballedo, Folgoso do Courel, Chantada, Bóveda, Saviñao, Sober, Pantón (Galicia). Complejo Hospitalario de Navarra: Lekunberri, Aoiz, Mendigorria, Sangüesa, Olite, Irurtzun, Puente de la Reina, Larraga, Falces (Navarre). Hospital Universitario Puerta del Mar: San Fernando, Cádiz (Andalusia). Complejo Hospitalario Universitario de Jaén: Martos, Jaén (Andalusia). Hospital Universitario Central de Asturias: Langreo, Oviedo (Asturias). Hospital de Barbastro: Barbastro, Monzón (Aragón). Hospital Comarcal de Inca: Inca, Sa Pobla (Balearic Islands). Hospital Universitario Nuestra Señora de la Candelaria: Güimar, Candelaria (Canary Islands). Hospital Universitario Marqués de Valdecilla: Camargo, Santander (Cantabria). Complejo Hospitalario Mancha Centro: Campo de Criptana, Alcázar de San Juan (Castilla La Mancha). Hospital Universitario Germans Trias i Pujol: El Masnou, Santa Coloma de Gramenet, Sant Adriá del Besós (Catalonia). Hospital Universitari Parc Taulí: Barberà de Vallès, Sabadell (Catalonia). Hospital de Mérida: Mérida, Almendralejo, Montijo (Extremadura). Fundación Hospital de Calahorra: Calahorra, Arnedo (La Rioja). Hospital Universitario Severo Ochoa: Leganés (Madrid). Hospital El Escorial: San Lorenzo de El Escorial, Galapagar (Madrid). Hospital Rafael Méndez: Águilas, Lorca (Murcia). Hospital Universitario de Cruces: Sestao, Baracaldo (Basque Country).

(0.72MB).

Interviewers in a call centre screened the sample using a telephonic questionnaire that covered two complementary routes: self-diagnosis and screening according to symptoms (Fig. 2, Appendix A, annex 1). If the patients stated that they had already been diagnosed, a rheumatologist subsequently confirmed their diagnosis by reviewing their clinical history. The patients who had no previous diagnosis, based on their replies to the symptom-based screening, either fulfilled the criteria for suspicion or did not. If they did not fulfil these criteria, their participation in the study was terminated. If they did fulfil the criteria, the next step consisted of a rheumatologist applying another more specific telephonic questionnaire (Appendix A, annex 2) to select the individuals with a persistent suspicion of diagnosis of arthrosis. These patients were given an appointment with the rheumatologist for a physical and radiological examination, to definitively confirm or rule out the diagnosis.

Fig. 2.

Call centre flow and telephonic questionnaire screening.

(0.22MB).

Cases were considered to have been lost when the subject completed the initial telephonic questionnaire with a positive result in the screening for arthrosis and the rheumatologist was unable to confirm or rule out the diagnosis.

All of the subjects were asked to give their informed consent orally in the first telephonic contact. This was also requested in writing from those participants who visited the surgery for physical examination and complementary tests. The study was approved by the Research Ethics Committee of the Hospital Universitario de Canarias, which acted as the reference committee, and it was also approved by the research ethics committees of the participating hospitals when this was required.

The prevalence of symptomatic arthrosis in Spain, in one or more of the locations studied (cervical and lumbar spinal column, hips, knees or hands), amounted to 29.35% (CI 95%: 27.77–30.97) and it increased with age, reaching its highest figures in subjects over the age of 80 years (52.6%; CI 95%: 46.97–58.29). It was more common in women, above all those over the age of 60 years (Fig. 3).

Fig. 3.

Prevalence of arthrosis in women and men.

(0.25MB).

The figures expressing the prevalence of arthrosis in the EPISER2016 study are generally higher than those corresponding to EPISER2000 (hands: 7.73% vs 6.2%; knees: 13.83% vs 10.2%). Nevertheless, when interpreting and comparing these results it is necessary to take into account the fact that the EPISER2016 study analysed the prevalence of arthrosis in the population ≥40 years old, while the EPISER2000 study did so for the population ≥20 years old.5

Another factor that has to be taken into account when comparing data is whether the studies defined cases as symptomatic or radiological arthrosis.2 Respecting this, the EPISER2000 study took symptomatic arthrosis into account for the hands and knees, as defined by ACR clinical criteria. In the EPISER2016 study, clinical and radiological criteria were used as well as the criteria used in EPISER2000 to diagnose knee arthrosis. The clinical criteria used in EPISER2000 were maintained for hand arthrosis.

An important novelty of EPISER2016 is the inclusion of spinal column arthrosis, giving rise to the first data on the prevalence of this condition in Spain. It is common in the population, in the cervical as well as the lumbar region. Few published works analyse the prevalence of axial arthrosis.16 Together with this lack of studies, there are also no validated criteria for the diagnosis of arthrosis in this location, so that we have been obliged to define them by combining clinical and radiological variables. These criteria were used to confirm suspicion in those cases that had not been diagnosed prior to the study.

EPISER2016 data support many of the known factors associated with arthrosis.17,18 Obesity is one of the most important of these factors. Many studies corroborate the association between obesity and arthrosis of the knees, hips and hands. The risk of suffering arthrosis is calculated to be 4 times higher in case of obesity for women and 4.8 times higher for men. EPISER2016 also shows an association with cervical and lumbar spinal column arthrosis. The causal mechanisms that would explain this associated have yet to be elucidated. A biomechanical cause would clearly not explain the association with hand or cervical spinal column arthrosis. The influence of metabolic factors in the pathogenesis of arthrosis is becoming increasingly relevant, underlining the existence of a metabolic phenotype for arthrosis.19

As is the case for obesity, sex and age are both factors which this study found to be associated with arthrosis in the 5 locations analysed. It occurs more frequently in women, and increased age seems to be more important factors for arthrosis of the lower limbs.

The association between tobacco consumption and arthrosis is under discussion: some studies, above all epidemiological ones, find tobacco use to be a protective factor, while others do not find any such association. In a 2017 systematic review and meta-analysis, the authors state that there is an inverse relationship between smoking cigarettes and the risk of knee arthrosis, chiefly in men. However, to date it has not been possible to find an explanation for this association, so that it has been suggested that more studies should be designed.20 Bivariant analysis in the EPISER2016 study shows a significant association between tobacco consumption and arthrosis in all of the locations studied, although this relationship does not emerge in the multivariant study.

The EPISER2016 study does not analyse the influence of occupation or work, although both factors were found by the NHANES-I and Framingham studies to contribute to the risk of arthrosis, above all of the knees, in professions that involve frequent use of the same, especially in bending (builders, firemen and sailors).17,18,21 Nevertheless, the EPISER2016 study includes educational level, and this may be understood to be a variable that is associated with occupation and work. The results show that a higher educational level is associated with reduced prevalence of axial as well as peripheral arthrosis, and not only in load-bearing joints such as the knees, hips and lumbar spinal column, but also in the hands and cervical spinal column.

The association between the geographical area where subjects live or were born and arthrosis has always been highly interesting. In this regard the most common finding is that it is more prevalent in rural areas and less developed countries.22 However, no statistically significant differences were found in our study between those who lived in a rural or urban area. Only the incidence of lumbar spinal column arthrosis was found to be higher in rural areas, although this was not statistically significant. On the contrary, knee arthrosis showed the opposite tendency; this finding is similar to that of the EPISER2000 study, which obtained an OR of 1.74 in multivariant analysis (CI 95%: 1.09–2.77), taking “rural” category as the reference.23

Bivariant statistical analysis of the variable “having been born abroad” was associated with cervical and knee arthrosis, although neither of these associations was maintained when adjusted for age (data not shown).

Cervical spinal column arthrosis was more frequent in the Mediterranean and central areas than it was in the north, while lumbar spinal column arthrosis was more frequent in the centre. Nevertheless, knee arthrosis seems to be more common in the Mediterranean area, with an association that is close to being statistically significant.

If we compare EPISER2016 data with those of publications in other countries, the prevalence of hand arthrosis can be seen to be similar to the rate published in the Framingham cohort of the U.S.A., while the incidence of hip arthrosis is similar to the population data published in the U.S.A., Portugal, Holland and France. On the other hand, the prevalence of symptomatic knee arthrosis is higher than the figures corresponding to the U.S.A., France and Norway, and it is more similar to the figures published in Portugal, Canada and Holland (Table 6).24–26 Comparative studies of the results obtained in each country always have to be approached with caution: the selection criteria for the population as well as the diagnostic criteria used are hard to standardise, and sometimes cause results to differ or be similar in ways that cannot solely be attributed to geographical distribution. This is why, to obtain more reliable results it is desirable that populations in different countries be analysed in a single study. A good alternative would be for the next EPISER study to be undertaken in collaboration with other scientific societies in different countries, to thereby obtain figures that can be compared.

Regarding the limitations of this study, it should be pointed out that in the oldest age group (those over the age of 80 years old) the percentage of individuals who mention being in poor or very poor health in the initial questionnaire was lower than the figure in the National Health Survey of 2017 (approximately 10%–15% lower; data not shown). This may have led the prevalence of arthrosis to be partially underestimated. Moreover, in the case of knee arthrosis, although the screening NPV was high (95.12%), it was slightly below the figure for the other locations, and this may have led to a certain degree of underestimation of its prevalence.

To summarise, the EPISER2016 study is the first one to analyse the prevalence of symptomatic arthrosis in 5 locations (the cervical and lumbar spinal column, knee, hips and hands) in Spain. The study shows that the overall prevalence of arthrosis in the population ≥40 years old stands at 29.35%. It also shows that the prevalence of hand and knee arthrosis is higher than it was 16 years ago (EPISER2000), although when comparing these results, it has to be taken into account that the selected populations were of different ages, and different diagnostic criteria were used. Of all the locations studied, lumbar spinal column arthrosis is the most prevalent.

EPISER2016 was financed by Celgene, Gebro Pharma, Merck Sharp and Dohme de Espana, Pfizer and Sanofi-Aventis. These financial backers did not intervene in the design of the study or data gathering and analysis, or in the writing of this paper. MS was financed by the Contrato Río Hortega-Fondo de Investigación Sanitaria (CM17/00101). The Scientific Research Fund, which is included in the National Scientific, Development and Technological Innovation Programme 2013–2016, is financed by the ISCIII-Subdirección General de Evaluación y Promoción de la Investigación- European Regional Development Fund (ERDF) “A way of making Europe”.

The authors have no conflict of interests to declare.