![[ Visit Client Website ]](images/banner.gif)
Purpose: Given the potential cardiovascular toxicity of serum urate, we sought to determine if treatment with allopurinol, a urate-lowering drug, is associated with lower mortality in hyperuricemic patients.
Methods: We conducted a retrospective cohort study using data from the Consumer Health Information and Performance Sets (CHIPS) database, which abstracts patient-level data from the Veterans Affairs (VA) electronic medical record system. The study population included veterans at least 40 years of age with hyperuricemia (outpatient serum uric acid level > 7.0 mg/dL) who received care from one of 8 VA facilities between 2000-2007. Hyperuricemic subjects filling their first prescription of allopurinol (exposed group) were matched to up to three hyperuricemic subjects not treated with allopurinol (unexposed group) and followed until death, study closure, or no further contact with the VA for 18 consecutive months. The association of allopurinol treatment with all-cause mortality was evaluated using Cox proportional hazards models, adjusting for potential confounders such as age, race, sex, baseline urate level, BMI, medications (especially cardiovascular medications such as diuretics), comorbid conditions (e.g. hypertension, diabetes, cardiovascular disease), and health care utilization.
Results: Among 9,924 veterans, there were 1,021 deaths during a median follow-up of 1.9 years. The cohort was comprised mostly of white males (87%) with a median age of 62 years. Compared to the unexposed group, the allopurinol treatment group had a higher prevalence of hypertension (81% vs 70%) and a higher mean serum urate level (9.0 mg/dL vs 8.2 md/dL) at baseline. After adjusting for baseline uric acid levels, allopurinol treatment was associated with a lower risk of all-cause mortality (hazard ratio [HR] 0.81; 95% confidence interval [CI] 0.70-0.94). Further adjustment did not appreciably alter this hazard estimate.
Table:
Conclusion: Allopurinol treatment is associated with a lower risk of all-cause mortality in hyperuricemic veterans, possibly reflecting a reduction in cardiovascular deaths. Any confounding-by-indication would be expected to bias the results towards the null, since the treatment group had a greater baseline cardiovascular risk. Replication of these findings in randomized controlled trials would permit more definitive conclusions regarding the role of urate-lowering drugs in the prevention of cardiovascular disease.
Methods: We conducted a retrospective cohort study using data from the Consumer Health Information and Performance Sets (CHIPS) database, which abstracts patient-level data from the Veterans Affairs (VA) electronic medical record system. The study population included veterans at least 40 years of age with hyperuricemia (outpatient serum uric acid level > 7.0 mg/dL) who received care from one of 8 VA facilities between 2000-2007. Hyperuricemic subjects filling their first prescription of allopurinol (exposed group) were matched to up to three hyperuricemic subjects not treated with allopurinol (unexposed group) and followed until death, study closure, or no further contact with the VA for 18 consecutive months. The association of allopurinol treatment with all-cause mortality was evaluated using Cox proportional hazards models, adjusting for potential confounders such as age, race, sex, baseline urate level, BMI, medications (especially cardiovascular medications such as diuretics), comorbid conditions (e.g. hypertension, diabetes, cardiovascular disease), and health care utilization.
Results: Among 9,924 veterans, there were 1,021 deaths during a median follow-up of 1.9 years. The cohort was comprised mostly of white males (87%) with a median age of 62 years. Compared to the unexposed group, the allopurinol treatment group had a higher prevalence of hypertension (81% vs 70%) and a higher mean serum urate level (9.0 mg/dL vs 8.2 md/dL) at baseline. After adjusting for baseline uric acid levels, allopurinol treatment was associated with a lower risk of all-cause mortality (hazard ratio [HR] 0.81; 95% confidence interval [CI] 0.70-0.94). Further adjustment did not appreciably alter this hazard estimate.
Table:
| Risk of Death by Allopurinol Use: Nested Adjusted Models (n=9,924) | ||
| MODEL | HAZARD RATIO | CONFIDENCE INTERVAL |
| Model 1: adjusted for baseline urate level | 0.81 | 0.70, 0.94 |
| Model 2: adjusted for baseline urate level, age, race & sex | 0.78 | 0.67, 0.90 |
| Model 3: same as Model 1, also adjusted for BMI | 0.79 | 0.68, 0.91 |
| Model 4: same as Model 2, also adjusted for comorbidities & health care utilization | 0.74 | 0.64, 0.86 |
| Model 5: same as Model 3, also adjusted for cardiovascular and other medications | 0.73 | 0.63, 0.85 |
| Model 6: same as Model 4, also adjusted for total cholesterol and GFR | 0.77 | 0.66, 0.90 |
Conclusion: Allopurinol treatment is associated with a lower risk of all-cause mortality in hyperuricemic veterans, possibly reflecting a reduction in cardiovascular deaths. Any confounding-by-indication would be expected to bias the results towards the null, since the treatment group had a greater baseline cardiovascular risk. Replication of these findings in randomized controlled trials would permit more definitive conclusions regarding the role of urate-lowering drugs in the prevention of cardiovascular disease.
A.J. Luk, None; E.E. Moore, TAP, 2; G.P. Levin, TAP, Abbott, 2; X. Zhou, None; H.K. Choi, None; B.R. Kestenbaum, None.