ROCHESTER, Minnesota—Rheumatoid arthritis (RA) patients are being diagnosed earlier and treated with more effective disease-modifying drugs but are living no longer today than they were 40 years ago. The result is a widening gap between life expectancy for RA patients and that of the general population, Mayo Clinic researchers will report in the November issue of Arthritis & Rheumatism.¹

"We found no evidence indicating that RA subjects experienced improvements in survival over the last 4 to 5 decades," said senior author Sherine Gabriel, MD. "In fact, RA subjects did not even experience the same improvements in survival as their peers without arthritis, resulting in a worsening of the relative mortality in more recent years, and a widening of the mortality gap between RA subjects and the general population throughout time." Dr. Gabriel is professor of medicine and of epidemiology at the Mayo Clinic in Rochester, Minnesota. She is also and an established NIH-funded investigator in the epidemiology of the rheumatic diseases.

These findings are likely to be disheartening for many who had hoped that earlier diagnosis, breakthrough drugs, and more aggressive treatment regimens were improving RA survival, or that people with RA were benefiting from the substantial and continuous increase in general life expectancy in the US and Europe during recent decades.

Population-based RA incidence cohort

Dr. Gabriel, lead author Angel Gonzalez, MD, and colleagues studied a population-based incidence Minnesota cohort of all Rochester residents >e;18 years who were first diagnosed with RA between 1955 and 1995 and all Olmsted County residents, in whom RA was first diagnosed between 1995 and 2000. They estimated expected mortality from the National Center for Health Statistics life tables using person-year methods.

The study cohort included 822 RA patients (72% women, mean age at RA incidence 58 years) followed for a median of 11.7 years, during which time 445 of them died. The researchers compared survival rates in five time periods: 1955 to 1964, 1965 to 1974, 1975 to 1984, 1985 to 1994, and 1995 to 2000 using Cox regression models, adjusting for age and sex.

"Between 1965 and 2005, the mortality rates across the calendar years for female and male RA patients were relatively constant at 2.4 and 2.5 per 100 person-years, respectively. In contrast, the expected mortality rate in the Minnesota white population decreased substantially over the same time period in both sexes," the authors write.

The researchers found that between 1965 and 2005

• mortality for female RA subjects remained at 2.4 per 100 person-years
• mortality for male RA subjects remained at 2.5 per 100 person-years
• mortality for women in the Minnesota general population declined from 1 per 100 person-years in 1965 to 0.2 per 100 person-years in 2000
• mortality for men in the Minnesota general population declined from 1.2 per 100 person-years in 1965 to 0.3 per 100 person-years in 2000

Estimated mortality at 10 years after diagnosis actually rose slightly for subjects with RA, from 24% for those diagnosed in 1955 to 1964, to 25% for those diagnosed in 1965 to 1974 and 1975 to 1984, and 29% for those diagnosed in 1985 to 1994 or 1995 to 2000.

The differences between mortality for RA patients and the general population were particularly stark for women. Mortality remained at 2.4 per 100 person-years for women with RA but dropped from 1 to 0.2 per 100 person-years for women in the general population. Mortality for men with RA remained at 25 per 100 person-years but dropped from 1.2 to 0.3 per 100 person-years for men in the general population.

Cardiovascular health advances not helping RA patients?

Much of the improved longevity in the general population has been due to declines in mortality from cardiovascular diseases. According to the researchers, "it is possible that the cardiovascular interventions that improved life expectancy in the general population may not have had the same beneficial effects in patients with RA."

The authors note that these findings may not be generalized to the nonwhite population, since the study population was predominantly white. Dr. Gonzalez also notes that data suggesting improved survival in recent years are likely skewed by survival bias, because survival was examined in prevalence cohorts rather than incidence cohorts.

Reference