VICTORIA, B.C. – The risk for cardiovascular disease is greater in people with osteoarthritis, but the reason why remains unclear, researchers reported at the annual meeting of the Canadian Rheumatology Association.
Using diagnosis codes, a team led by M. Mushfiqur Rahman, a biostatistician at the Arthritis Research Center of the University of British Columbia, Vancouver and a doctoral student with the School of Population and Public Health there, compared risk by osteoarthritis (OA) in a random sample of adults from the province’s general population who were free of cardiovascular disease (CVD) at baseline.
M. Mushfiqur Rahman
In all, 12,624 people with prevalent or incident OA were matched with 61,131 unaffected people. With a mean follow-up of 13 years, there were 1.3 new cases of CVD per 100 person-years in the study sample, according to results reported in a poster session.
Analyses showed a statistically significant 19% increase in the risk of CVD in people who had OA vs. those without, after multivariate adjustment for a variety of potential confounders, such as hypertension, diabetes, and comorbidity burden.
In sex-stratified analyses, the elevation of risk was below average for men, whether they were younger than 65 years of age (12% increase in risk) or older than this age (13%). But it was notably above average for both women younger than age 65 years (41%) and women over this age (27%).
The mechanisms underlying the observed association are unknown, according to Mr. Rahman. Inflammation is one possibility, although there is no evidence to support this relationship at present.
"We cannot explain this relationship biologically," he commented in an interview. "Maybe reduced physical activity is a major factor here: After a diagnosis of OA, people are not as physically active as they were before."
"Prescription medication could also be a big factor," Mr. Rahman continued. "We didn’t adjust for any prescription drugs here," although patients with [OA] are more likely to be using acetaminophen, nonsteroidal anti-inflammatory drugs, and opioids. "Those medications may have an effect [in elevating CVD risk]. We will see in the future."
The investigators are obtaining linked self-reported data that may shed further light on mechanisms. They will also be cross-checking the findings by assessing CVD mortality in patients with OA. If patients with OA prove to have higher CVD mortality, "that will be a good validation study."
In the meantime, physicians who see patients with OA may want to counsel them, "do more physical activity and take care of your heart," said Mr. Rahman.
Study participants were a random sample of individuals drawn from Population Data British Columbia for the years 1991 to 2009.
They were defined as having OA if the OA diagnostic code was used on either two or more visits to a health professional in 2 years or at least one discharge from the hospital. Similarly, they were defined as having CVD if they had a hospital record having a diagnosis code for ischemic heart disease, congestive heart failure, or stroke.
The mean age was about 58 years, and 59% of participants were women. About 16% had hypertension, 5% had hyperlipidemia, and 2% were obese. The mean Charlson comorbidity index was 0.41 in the OA group and 0.32 in the control group.
In addition to the elevated risk of CVD associated with OA (relative risk, 1.19), study results showed that participants were more likely to receive a CVD diagnosis if they had chronic obstructive pulmonary disease (1.18), had hypertension (1.38), were obese (1.21), or were in the lowest quintile of socioeconomic status (1.05). Also, risk increased with Charlson comorbidity index (1.11).
"There are some unmeasured confounders, for sure," Mr. Rahman acknowledged, noting that some variables, prescription drug use among them, were not available in the database used. Obesity was assessed from physician diagnoses instead of from body mass index; thus, the prevalence may have been underestimated. Also, "although we adjusted for [chronic obstructive pulmonary disease], smoking is an unmeasured confounder. And the physical activity level of the patients could be an unmeasured confounder, too."
Mr. Rahman disclosed that he had no relevant conflicts of interest.