Rheumatoid arthritis (RA) is a relatively common autoimmune disease that causes progressive health impairment. Its prevalence in the general population of Spain is 0.5%,1 whereas the number of new cases arising annually is approximately 8.3 cases per 100,000 people (95% confidence interval (CI): 7.5–9.2).2 RA not only impairs the musculoskeletal system but can also affect the entire body and increase morbidity and mortality. Both age and physical disability affect health-related quality of life and the costs associated with managing the disease.3 The coincidence of two or more diseases in a patient with RA, a phenomenon known as multimorbidity, worsens the health-related quality of life as well as the major health outcomes of RA, including the costs of coexisting disease.4 In fact, the rate of disability increases with the number of comorbid diseases.5 Moreover, the number of comorbidities is an independent risk factor for premature death6 and drug toxicity7 in RA patients.

Because early and aggressive control of the inflammation associated with RA results in improved survival,8,9 the addition of new biological agents in the treatment of RA could help to reduce mortality. However, fewer than 30% of patients with RA currently take biological agents.7 The patients who receive anti-TNF therapies often have the most severe disease, with high levels of disability. However, patients with high levels of comorbidity are rarely prescribed anti-TNF therapy but may remain on standard DMARD therapies alone.There are few studies that analyse comorbidities in patients with RA who will receive biological therapy. Furthermore, there are no specific studies employing the design used in our study to compare comorbidities in patients with or without biological drugs. As RA is a prevalent disease, it is important to consider comorbidities; in our opinion, it is critical to impart these findings to the medical society.

Our aim was to determine whether RA patients who have been prescribed biological agents to control their inflammatory activity have a comorbidity burden different from those who were prescribed disease-modifying antirheumatic drugs (DMARDs) and to understand the associations between comorbidity and other variables and between comorbidity and multimorbidity.

This observational case–control study was performed in a tertiary care centre (reference population of 628,912 inhabitants; current annual load of approximately 1500 follow-up RA patients). All the patients who entered the cohort signed a written consent form after being informed regarding the details of the study, according to the Declaration of Helsinki and Spanish research regulations. The study protocol was also reviewed and approved by the centre's ethical committee.

The clinical status of the cases and controls is shown in Table 1. The mean age (SD) of the RA patients (79% women) was 54.6 (11.8) years (range, 16–80; median 56), and the mean disease duration was 7.9 (7.7) years.

Compared with patients who were prescribed only DMARDs, the cases had the disease for a longer time and exhibited a higher frequency of subcutaneous nodules, an increased inflammatory activity, and worse physical function. These patients had also previously received a greater number of DMARDs, and a higher proportion was not currently using DMARDs. Furthermore, more than half of the study groups were receiving corticosteroids. Total joint arthroplasties were twice more common in the RA patients requiring biological therapy than in the patients taking only DMARDs. Only certain patients who were not prescribed biological agents received combination DMARD therapy. Other characteristics did not differ between the groups.

We have analysed and compared the relative frequency of comorbidity and multimorbidity in two sex- and age-matched populations of RA patients. Our study group comprised RA patients treated with biological therapy, and our control group comprised RA patients treated with DMARDs alone. Together, these subjects constitute a representative sample of patients with RA, and their clinical and epidemiological characteristics are similar to those found in most large hospital series22 or biological registries.23

RA patients who are prescribed biological agents often have a severe form of the disease and a worse clinical course.24 The observations of our study group support this finding. We found that patients who were prescribed biological agents had higher HAQ scores, had previously received more DMARDs, used more glucocorticoids, and had a higher frequency of total joint arthroplasties than the patients in the control group. Although the elevation of the DAS28 could be explained by the manner in which the patients were selected, a mean baseline DAS28 of 6 indicates severe disease. Despite baseline differences, our study indicates that the relative frequency of comorbidities was similar in the patients who were prescribed biological therapy and those RA patients who were prescribed only DMARDs. This lack of association between inflammatory severity and comorbidity has been observed previously.25 These authors studied a community-based cohort of 183 patients in the early stages of RA, and they observed that measures of inflammation at diagnosis or during follow-up were not predictive of the development of cardiovascular disease. Moreover, they also observed that the vast majority of all comorbidities occurred more than 10 years after the onset of RA, whereas the mean duration of arthritis in our cohort was less than 10 years. Although our univariate analyses revealed that activity factors were associated with comorbidity and multimorbidity, our multivariate analyses revealed no such association. However, this discrepancy may be due to the small sample size or may be because the activity we measured (DAS28, ESR, and CRP) did not account for the accumulated activity burden over time.

Although comorbidity levels in patients treated with biological agents have been studied previously,7 ours is the only study with a control for sex and age. Information regarding the association between the biological treatment of RA and the comorbidity profile of patients is medically relevant and can be of prognostic value. Additionally, the more comorbid illnesses one has, the greater is the interference with treatment, leading to greater medical costs, disability and risk of mortality.4 The most frequent comorbidities in our patients were hypertension, obesity, diabetes, and respiratory, thyroid, and upper gastrointestinal disorders. The relative frequencies of these comorbidities are equivalent to those of other studies. Hypercholesterolaemia was the most prevalent cardiovascular risk factor and exhibited a negative association with the need for biological therapy. Although this difference did not reach statistical significance, higher disease activity is associated with lower total cholesterol levels26 and significantly depressed high-density lipoprotein levels, leading to a higher atherogenic index.27

A limitation of our study is its retrospective design. In a case-controlled study such as ours, the investigator must work backwards. Most comorbidity studies use retrospective cohorts and cross-sectional studies in which data concerning comorbid illnesses are retrieved from patient self-reporting or their medical records. We used data reported by patients and data obtained from medical records. Moreover, both groups are comparable because they were selected consecutively and were matched with respect to the two variables with the greatest impact on the burden of comorbidity, age, and sex.28

Our results are consistent with previous studies and show that the presence of comorbidities and cardiovascular risk factors affects the majority of patients with RA regardless of biological therapy.6,7,29–31 It is well known that in comparison with other chronic diseases, patients with RA suffer a disproportionate comorbidity burden. Similarly high comorbidity is associated with osteoarthritis,6,32 but high comorbidity is not associated with all rheumatic diseases.33 However, the methods used by these studies vary greatly and therefore, the results are difficult to compare. One of the main differences between studies involves the types of diseases and/or disorders included as comorbidities. Associated conditions are diverse and can range from a simple cataract to terminal cancer. We mainly considered the conditions listed in the Charlson index21 (based on 17 diagnoses, each weighted by mortality risks) extended with a few chronic diseases that emphasise measures of impairment and disability. In a disease such as RA, it is important to consider the comorbidities related to mortality and disability outcomes.34

Considerable evidence indicates that patients with RA are at greater risk for developing coronary heart disease.35–37 Although we acknowledge that there may be an underestimation, only 2% of our patients had cardiovascular disease. This figure is lower than that observed by Gabriel et al.6 but the patients in their study were 10 years older because the study included three prevalence cohorts (from 1965) assembled with RA patients that were greater than 63 years old. However, cardiovascular disease is less common in women under 60 years, and our sample consisted of 80% women, 65% of whom were under 60 years. The prevalence of cardiovascular disease in our study is comparable with the latest reports on RA patients treated with biological drugs in this age group.38

Although we know that the risk of multimorbidity in RA is high, we do not understand what heightens this risk. It is possible that RA and its comorbid conditions share the same risk factors. For example, smoking increases the risk of cardiovascular diseases, cancer, and RA. In our study, we identified four factors associated with comorbidity in the univariate analysis, two related to age and the rest related to the severity of RA. However, age was the only predictive factor for both comorbidity and multimorbidity in the multivariate analyses. Because many health problems are known to increase with age, this demographic trend may lead to an increase in the absolute number of chronic conditions in RA patients. For this reason, age always appears as the strongest predictor of comorbidity in addition to pre-existing comorbidity in RA patients.25,39,40

In conclusion, RA patients who are prescribed biological drugs exhibit a comorbidity burden equivalent to those treated with DMARDs only. Age is the main predictive factor of comorbidity in these patients.

Dr. Antonio Fernández Nebro: papers and advice for ABBOT, MSD, PFIZER, ROCHE.