Rheumatic and musculoskeletal diseases are among the most common non-transmittable chronic diseases.1 Among these, Osteoarthritis (OA) is the most frequent in young adults and older individuals and is associated with short and long-term disability.2,3 OA is characterized by joint pain, stiffness, inflammation, and weakness, which lead to physical disability.4

In Latin America, by using the COPCORD methodology, OA prevalence varies between 2.3% up to 20.4%; primary or idiopathic OA represents 88.2% of cases,5 with the knee being the most affected (51.1%) and associated with a Health Assessment Questionnaire (HAQ)>0.375 (OR 1.9, 95% CI 1.5–2.4); in addition, the inability to kneel is a predictor for patients seeking healthcare visits.6

Worldwide, OA is the fourth most frequent cause of years lived with disability (YLDs),7 whereas in Mexico it ranks 8th among women an 13th in men.8 Some studies report that up to 50% of patients with symptomatic OA suffer from some degree of disability.9 According to the World Health Organization (WHO), “disability” is a general term covering impairments, activity limitations, and participation restrictions. An impairment is a problem in body function or structure; an activity limitation is a difficulty encountered by an individual in executing a task or action; while participation restriction is a problem experienced by an individual I involvement in life situations.10

Disability increases with age and disease duration, with pain being one of the most important causes of disability in patients with OA9,11–13. It has been reported that OA occurs in at least 10% of adults 60 years and older, it can damage any joint, with knee and hip the most affected.14

According to several studies, the main co -morbidities that influence OA are: obesity, diabetes mellitus, hypertension, dyslipidemia, metabolic syndrome, gout, and pulmonary diseases,5,14–18 with obesity and diabetes being the most associated with OA and the presence of disability.16–19

In Mexico, patient care with rheumatic and musculoskeletal diseases can vary widely since there a very peculiar health care system, and healthcare coverage is based on patient emplo yment and economic income.20

The aim of this study is to determine the association of co -morbidities and their impact on disability, since it represents the most important outcome at the public health level.

Transverse, analytic and multicenter study of the “Iniciativa Mexicana de pacientes con Osteoartritis y Artritis Reumatoide” (IMPACTAR), in a nationwide sample involving 55 centers in 20 states in Mexico.

This study included 4971 consecutive patients 18 years and older with the diagnosis of OA according to American College of Rheumatology (ACR) classification criteria that received ambulatory care in public, private or social security institutions across Mexico during 2013–2015.

Patients belonged to different federative entities or states from the whole country, in order to classify their origin, we used the socio-economic region classification from the “Instituto Nacional de Estadística y Geografía” (INEGI), an autonomous agency of the Mexican Government to coordinate the National System of Statistical and Geographical Information.21

Prior informed consent, each subject completed a questionnaire regarding socioeconomic, clinical, healthcare, and physical disability aspects.

Disability was evaluated using the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), which was designed to measure functional capacity in patients with OA and contains 24 items grouped in 3 different domains: pain (0–20), stiffness (0–8) and functional limitation (0–68); each item having 5 levels which are codified from 0 to 4 (none, mild, moderate, severe, and extreme).22,23 These domains result in an index score ranging from 0 to 12, and from these, two categories were created: absence of disability (<3) and presence of disability (> or =3).24

The IMPACTAR registry accrued a total of 7073 patients; of these 4971 were selected by having a diagnosis of OA and a complete functional evaluation (WOMAC). Mean age was 63 years, 75% were female, with most having a basic education level (49.9%). Two groups were created: One with presence of disability (41.5%) and the other without (58.5%). (Table 1)

In this study most subjects belonged to region 7 which corresponds to Mexico City, the states of Nuevo León, Coahuila and Jalisco with 28.9% followed by region 5 comprising the states of Baja California Norte, Baja California, Sonora, Chihuahua and Tamaulipas with 21.7%, whereas the least represented region was region 1 which corresponds to the states of Guerrero, Oaxaca, and Chiapas with 1.4%.

The median family income in our cohort was $13 000 MXP (IQR: 9800–16 000) (Table 1).

The median Body Mass Index (BMI) was 28.4 (26.31.6) kg/m2, and according to the WHO classification, 86.7% of patients were obese or overweight.

The general median to the time of OA diagnosis was 4 years (IQR 1–7 years), and 51.2% of patients had less than 5 years since diagnosis. The most affected joint in OA patients was the knee (88.8%), and up to 13.5% of patients reported history of joint surgery.

When we investigated the different types of medications taken by patients at the time of the interview, we found that 48.1% took non-steroidal anti-inflammatory drugs (NSAIDs).

More than 49% of subjects had at least one co-morbidity, with hypertension being the most prevalent in 32.8% (Table 2)

Regarding the type of health care, the patients belonged to, 14.6% belonged to public institutions, 28.4% to social security and the rest to private medical practice.

The strength of this study is that it offers a global view of the current situation regarding OA patients when being attended by internal medicine, rheumatologists, and orthopedics specialists, whereas other studies exclude these patients or include only those being taken care of by primary care physicians. Furthermore, this analysis comes from a national registry of almost 5000 OA patients, the largest published so far. Our study reflects the characteristics of 4971 OA patients cataloged by 7 socioeconomic regions and different healthcare types and assesses the impact of co-morbidities toward disability through the WOMAC questionnaire. Regarding patient age and gender, our cohort showed similar results to other previously published studies.

Regarding the healthcare professionals that participated in the IMPACTAR study, we presented significant differences with studies previously done in Mexico,6,12 while Burgos et al. and the PANLAR study reported patients evaluated only by rheumatologists, the IMPACTAR initiative included patients evaluated by physicians from all medical specialties and subspecialties related to the management and treatment of OA which allows for a wider healthcare perspective.12

In our transversal study, 75% of our sample was comprised by women,25 which is in accordance with other studies of knee OA. It is also in accordance with community -based reports using the COPCORD methodology where the mean OA prevalence in Mexico is 10.5%, although some northern regions like Nuevo León reported 17.3% with also, female predominance. These results coincide with those found by Burgos et al., and others.5,12,26

Regarding the most affected anatomical site in different cohorts, in the IMPACTAR registry, 88% was the knee, a similar proportion reported by Burgos et al.,12 whereas in the PANLAR study the knee represented only 31%. These differences may be explained by the wider population diversity in the PANLAR study. Furthermore, there may be selection, diagnosis and referral bias, since due to its clinical characteristics, knee OA is easier to diagnose, and frequently requires orthopedic and rehabilitation evaluations and less care by rheumatologists, especially if it has already caused disability.

Another contrasting finding was disease latency; Burgos et al.12 reported a mean duration of 69 months, while in patients in the IMPACTAR registry was 48 months; this may be due to the heterogeneity of medical specialists that took care of patients in our cohort.

It is also worth noting that in the study by Burgos et al.,12 one quarter of patients had a low WOMAC score and minimal physical limitations despite the diagnosis. In our study, we observed that a larger number of patients (41.5%) from different regions had disability as measured by a WOMAC above or equal to 3, which reflects the type of population that is seen by medical specialists.

Previous OA cohorts in Mexico had not considered adjusting variables like socioeconomic income, education level, socioeconomic region or biochemical profile.6 We registered these variables and they were evaluated in the statistical model.

Some authors have found an association with two co -morbidities in patients with knee OA.25 In our study, 49.6% of patients had at least one comorbidity. We report a high prevalence of hypertension (HTN) (32.8%) in our population. Barquera et al.,27 in the 2006 national health survey, found an HTN prevalence of 31.5% (CI 95% 29.8–33.1) which coincides with our results.

A study done in Cuba by Friol GJ et al., found that half of patients referred co-morbidities associated with OA, with diabetes mellitus and obesity being the most frequent, either alone or combined; obesity was present in 13.5% of subjects and diabetes mellitus in 9%, even though almost 75% were above 60 years old.9

It is expected that when Body Mass Index (BMI) increases, the number of co-morbidities increases and physical activity decreases.28 In our study, average BMI was 28.3 (26.3–31.6) in patients with OA without disability and there was no significant difference to those who had disability.

The presence of disability was defined as a WOMAC function score ≥3 and was analyzed in a dichotomous manner. It is possible that the cutoff point does not discriminate adequately patients with different types of co-morbidities due to the low quantity of subjects with one severe disability for this cutoff point in the WOMAC scale. Nonetheless, the definition of disability was mostly designed toward detecting patients without disability.

The economic income in patients with disability was lower ($13 000 MXP) (IQR: 9000–16 000) adjusted to relevant co-variables like gender, age, time from diagnosis, co-morbidity and education level. This agrees with previously published data in other studies of musculoskeletal pain where socioeconomic privation has been described as a risk factor for pain in general as well as chronic pain.26,29,30 Also, the WOMAC pain dominion has been associated with poverty level. Data indicate that disability leads to poverty and poverty is associated with greater disability.29 We observed a statistically significant difference between patients with or without disability according to the type of institution they received care from, where the group with disability who were taken care for in public hospitals (16.1%) was higher in comparison to those without disability (13.6%), and this may be bound to the former.

We found that 48.1% of patients took NSAIDs, which is very similar to that observed in community-based studies where NSAIDs are the most frequently used medications (up to 60%).5,26 Moreover, we observed that patients who reported NSAID intake were more likely to be disabled (50.3%) which would reflect in a higher NSAID associated co-morbidity; however, we were unable to establish this association in this study.

Age, education level and time from OA diagnosis were not significantly associated with the presence of disability under an adjusted model, and this may be due to a large proportion of patients having a low education level. In Mexico, we still have low education levels, and in this study, we observed that half of OA patients had only basic schooling level (49.9%). Cleveland et al, found that the WOMAC functional domain and total scores were associated with schooling level, where OA patients with low education levels had more pain and worse function.

We report a significant nonparametric correlation (Rho=0.82) between the WOMAC pain and function dominions.

It is important to understand the disability process and characterize the changes that occur during its development and establish when it is reaching a critical point in order to incorporate preventive strategies; ideally, this would be during preclinical asymptomatic stages or early asymptomatic radiographic stages.

There are weaknesses in our study. One was that we did not evaluate the radiographic OA stage with disability nor the presence of comorbidities. Also, we only included patients that attend to hospitals or medical specialists and thus, it will not reflect the state of disability or presence of co -morbidities in the open population. We could not evaluate nor compare in our study if patients with OA and disability had more hospital admissions, out of pocket expenses, or if comorbidity decompensation resulted in more hospitalizations or physician office visits per year. We were unable to determine if the increased use of these medical services is due specifically to OA or other potential confounding factors.

Patients with OA and presence of co-morbidity have almost a 50% more chance of having disability than those without co-morbidities. Regions in Mexico show different proportions of patients with disability, and a higher economic income was associated with a decreased risk of having disability. There is a need for programs and interventions to decrease disability in OA patients while taking in consideration co-morbidities, female gender, family economic income and geographic region origin as variables that are associated to disability in OA.

Colegio Mexicano de Ortopedia y Traumatología, through an unrestricted educational grant from the pharmacological company SANOFI. SANOFI did not participate in the elaboration of this manuscript or in data analysis presented.

Data confidentiality: The authors declare that they have followed the established protocols in their respective institutions in regards to patient data confidentiality, and that they have followed ethical guidelines or rules to publish with patient informed consent, and that all patients included in this study were provided with sufficient information about this study and gave their informed consent to participate.

All authors declare to have no conflicts of interest.