Background and purpose
Serum uric acid (SUA) is a biomarker that has drawn attention for potential links to cardiovascular disease and mortality. A population-based prospective cohort study using data from the Berlin Initiative Study (BIS) examined whether SUA levels in community-dwelling older adults influence the risk of major adverse cardiovascular events (MACE) and death. The BIS started in 2009 to investigate chronic kidney disease among people aged 70 and older, providing a rich data source with repeated SUA measurements and linked health claims data.
Study population and design
The BIS included 2,069 initially eligible participants aged 70+ who were members of a statutory health insurance fund in Germany. After applying exclusion criteria and ensuring valid SUA assessments with longitudinal linkage, 2,058 participants contributed data for analysis. Participants were followed from baseline until an outcome occurred, up to four years after the last SUA measurement or the study end in 2021. SUA was measured with a standardized enzymatic colorimetric method and updated biennially, enabling a time-varying exposure approach.
Exposure, outcomes and covariates
SUA values were categorized into tertiles based on baseline distribution: lower (1.68–5.16 mg/dl), intermediate (5.17–6.48 mg/dl), and higher (6.49–13.00 mg/dl). SUA values were treated as time-varying and updated at each visit, with missing values imputed using the last observation carried forward. The primary outcomes were MACE (non-fatal stroke, non-fatal myocardial infarction, and cardiovascular death) and all-cause mortality. Outcomes leveraged a combination of self-reports validated against hospital records and death certificates; cardiovascular deaths were adjudicated by two physicians.
Covariates included age (modeled with splines), sex, education level, body mass index, smoking, alcohol use, physical activity, estimated glomerular filtration rate (eGFR) using BIS2, urine albumin-to-creatinine ratio (ACR), and histories of hypertension, diabetes, hyperlipidemia, cancer, prior MI/stroke, heart failure, and peripheral vascular disease. SUA-related medications were recorded but not treated as confounders due to lack of observed association with outcomes.
Statistical approach and sensitivity analyses
Hazard ratios (HRs) for the risk of MACE and all-cause mortality were estimated using Cox models with time-fixed covariates and SUA as a time-varying exposure. Intermediate SUA served as the reference. The team conducted several sensitivity analyses to account for time-dependent confounding, including models adjusting for eGFR and ACR as time-varying factors and marginal structural models with inverse probability weighting. They also performed post-hoc analyses using standard reference ranges for SUA and explored non-linear associations with flexible modeling (splines).
Key findings
Across the BIS cohort, higher SUA levels were associated with a higher risk of all-cause mortality compared with intermediate SUA levels in initial analyses (adjusted HR around 1.26). However, this association attenuated and largely disappeared when time-dependent confounding by kidney function (eGFR) and kidney damage (ACR) was accounted for, particularly in marginal structural models. In contrast, SUA levels did not show a robust association with MACE after adjustments. Secondary analyses showed no major effect modification by diabetes status or sex; an age-related pattern emerged where higher SUA appeared more strongly linked to mortality among septuagenarians (70–<80 years) than octogenarians, though interaction results were modest.
Interpretation and implications
The BIS findings challenge the notion that SUA is a straightforward predictor of cardiovascular events in very old adults. While higher SUA was initially linked to higher all-cause mortality, this signal largely disappeared after accounting for time-varying kidney function and damage, suggesting that SUA may be a marker of underlying comorbidity rather than a causal driver. The study reinforces the importance of considering time-dependent factors and frailty when assessing biomarkers in aging populations.
From a clinical perspective, the results do not support routine treatment of asymptomatic hyperuricemia in older adults solely to reduce MACE or mortality risk. This aligns with current guidelines cautioning against SUA-lowering therapy in asymptomatic individuals and highlights the need to weigh potential medication risks in polypharmacy-rich geriatric care.
Strengths and limitations
Key strengths include the time-varying SUA exposure and robust, multi-source outcome adjudication within a well-defined older population, enhancing external validity for similar settings. Limitations involve residual confounding inherent to observational data and the complexity of teasing apart SUA’s role from kidney function and age-related health factors. Some secondary analyses yielded imprecise estimates, underscoring the exploratory nature of those findings.
Conclusion
In this older German cohort, higher SUA levels were not consistently linked to increased MACE risk, and any association with all-cause mortality diminished when time-dependent kidney function was considered. The results suggest SUA alone should not drive treatment decisions in asymptomatic older adults and emphasize the broader context of kidney health and aging when interpreting biomarker data.
