Article Text

Download PDFPDF

Concise report
Relationship between serum urate concentration and clinically evident incident gout: an individual participant data analysis
  1. Nicola Dalbeth1,
  2. Amanda Phipps-Green2,
  3. Christopher Frampton3,
  4. Tuhina Neogi4,
  5. William J Taylor5,
  6. Tony R Merriman2
  1. 1 Department of Medicine, University of Auckland, Auckland, New Zealand
  2. 2 Department of Biochemistry, University of Otago, Dunedin, New Zealand
  3. 3 Department of Medicine, University of Otago, Christchurch, Christchurch, New Zealand
  4. 4 Clinical Epidemiology Research & Training Unit, Boston University School of Medicine, Boston, Massachusetts, USA
  5. 5 Department of Medicine, University of Otago, Wellington, New Zealand
  1. Correspondence to Prof Nicola Dalbeth, Department of Medicine, Faculty of Medical and Health Sciences, University of Auckland, Auckland 1, New Zealand; n.dalbeth{at}auckland.ac.nz

Footnotes

  • Handling editor Tore K Kvien

  • Contributors ND conceived and planned the study, interpreted the data, and wrote the first draft of the manuscript. ND is responsible for the overall content of the manuscript as guarantor. AP-G contributed to analysis planning, data management and manuscript drafting. CF analysed the data and contributed to analysis planning and manuscript drafting. TN, WJT and TRM contributed to study conception and planning, data interpretation, and manuscript drafting.

  • Funding This work was supported by the Health Research Council of New Zealand. TN is supported by NIH K24 AR070892.

  • Competing interests ND has received consulting fees, grants or speaker fees from Takeda, Horizon, Menarini, AstraZeneca, Ardea, Pfizer, Amgen and Kowa outside of the submitted work. WJT has received consulting fees from AstraZeneca and Pfizer outside of the submitted work. TRM has received consulting fees or grants from Horizon, AstraZeneca and Ardea outside of the submitted work. The other authors have no financial disclosures.

  • Patient consent Analysis of publicly available data.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Correction notice This article has been corrected since it published Online First. The results section of the abstract has been corrected.

  • Presented at This manuscript is based on work previously presented at the American College of Rheumatology Annual Scientific Meeting, San Francisco, USA, 2017, and was published as the following conference abstract: Dalbeth N, Phipps-Green A, Frampton C, Neogi T, Taylor WJ, Merriman TR. The Relationship between Serum Urate Concentration and Incident Gout: An Individual Participant Data Analysis [abstract]. Arthritis Rheumatol. 2017; 69 (suppl 10).

View Full Text

Statistics from Altmetric.com

Elevated serum urate concentration (hyperuricaemia) is considered to be a key risk factor for developing gout.1 However, a wide variation in estimates of risk has been reported depending on different published studies.2–5 The prevalence of hyperuricaemia is increasing.6 Understanding the consequences and risks of hyperuricaemia is important, both from a health planning perspective and to inform advice given to individuals with hyperuricaemia. The aim of this analysis of individual participant data was to examine the relationship between serum urate concentration and clinically evident incident gout, and specifically to provide estimates of the cumulative incidence of gout according to baseline serum urate concentrations.

Methods

The eligibility criteria for inclusion in this individual participant data analysis were any longitudinal observational study with the following data: publicly available patient-level data; incident gout data available; gout defined using recognised classification criteria, or doctor diagnosis, or patient self-report of disease, or self-report of doctor diagnosis; serum urate measured prior to assessment for incident gout; and a minimum duration of 3 years of follow-up.

Studies were identified through a systematic search of PubMed and the Database of Genotype and Phenotype (DbGaP) on 11 June 2016, searching the following terms: ‘gout and uric’. All articles published since 1 January 1980 were reviewed. The reference lists from comprehensive reviews and identified prospective studies were manually searched. The results of the various searches were reviewed by two authors (ND and TRM). Individual participant-level data were accessed from DbGaP (Project #834: The genetic basis of gout). Full details about data extraction are shown in online supplementary methods.

Supplementary file 1

For all cohorts, the baseline visit was selected as the first examination with both serum urate-specific and gout-specific data publicly available. The cumulative incidence of gout was calculated for the entire group per mg/dL serum urate category (primary analysis), and then for men and women separately (secondary analysis), over 3 years, 5 years, 10 years and 15 years. The incidence results were summarised as both cumulative incidence (%) and as the number of incident cases per 1000 person-years.

The time to onset of gout for each stratum of serum urate at baseline was examined using Kaplan-Meier estimates. Cox proportional hazard multivariable modelling was employed to model the hazard of gout for each category of serum urate at baseline (compared with <6 mg/dL); this analysis included age, sex, ethnicity and cohort as covariates. Where available, the date of gout diagnosis was used in the primary analysis, and if not available examination date at which gout was first reported. In sensitivity analyses, the examination date at which incident gout was recorded and the interaction term of cohort × baseline serum urate were included in the models. No imputation for missing data was undertaken. All analyses were undertaken using the Statistical Package for Social Sciences (SPSS) V.24.0 software.

Results

Search results

Search results are shown in online supplementary figure 1. Four cohorts with publicly available data fulfilled the inclusion criteria and were included in this analysis: Atherosclerosis Risk in Communities Study (ARIC), Coronary Artery Risk Development in Young Adults Study (CARDIA), and both the Original and Offspring cohorts of the Framingham Heart Study (FHS).

Supplementary file 2

Cohort and participant characteristics

Full methods and characteristics of the cohorts have been reported previously.7–10 Detailed cohort and participant details included in this analysis are shown in online supplementary table 1. For both ARIC and CARDIA, the diagnosis date of incident gout was available. For the two FHS cohorts, only the examination date at which incident gout was recorded was available. There were a total of 18 889 participants who were gout-free at baseline, with mean (SD) 11.2 (4.2) years and 212 363 total patient-years of follow-up. Overall, there were 8280 men (43.8%) in the analysis, and the mean (range) at the time of baseline serum urate testing was 49 (17–85) years.

Supplementary file 3

The ARIC cohort contributed the largest number of participants (n=10 775), followed by CARDIA (n=3470). Similar percentages of men were included in all cohorts (42%–46%). There were differences in age at baseline between the cohorts, with younger participants in CARDIA compared with other cohorts. Ethnicity also differed between cohorts, with predominantly white participants in the FHS cohorts, and both white and African–American participants in the ARIC and CARDIA cohorts. Serum urate concentrations at baseline were higher in the ARIC cohort, compared with the other cohorts.

Cumulative incidence of gout

For all participants, the overall cumulative incidence (95% CI) of gout by 3 years was 0.6% (0.4 to 0.8), by 5 years was 1.1% (0.9 to 1.3), by 10 years was 2.4% (2.2 to 2.6) and by 15 years was 3.2% (2.8 to 3.6) (table 1). The cumulative incidence of gout was lower in women than in men for all time points (table 1).

The cumulative incidence for each time point increased in a non-linear concentration-dependent manner according to baseline serum urate concentration (table 1, figure 1, online supplementary figure 2). By 5 years, the cumulative incidence (95% CI) ranged from 0.33% (0.23 to 0.43) for baseline serum urate <6 mg/dL to 26% (17 to 36) for ≥10 mg/dL (table 1, online supplementary table 2). By 10 years, the cumulative incidence (95% CI) ranged from 0.79% (0.63 to 0.96) for baseline serum urate <6 mg/dL to 40% (29 to 51) for ≥10 mg/dL. The 15-year cumulative incidence (95% CI) ranged from 1.1% (0.90 to 1.4) for baseline serum urate <6 mg/dL to 49% (31 to 67) for ≥10 mg/dL.

Supplementary file 4

Table 1

Cumulative incidence based on baseline serum urate groups for the entire group, men and women for each follow-up period

Figure 1

Kaplan-Meier plot showing the percentage of participants who were gout-free over the follow-up period, based on baseline serum urate categories in mg/dL.

Based on the 5-year data, the incidence of gout for those with serum urate ≥7 mg/dL was 9.8/1000 person-years, for ≥8 mg/dL was 20/1000 person-years, for ≥9 mg/dL was 34/1000 person-years and ≥10 mg/dL was 53/1000 person-years (online supplementary tables 3–4).

For women with high serum urate concentrations at baseline, the cumulative incidence of gout was lower by 3 years of follow-up compared with men with the equivalent high serum urate concentrations at baseline (table 1 and online supplementary figure 2). However, for the longer time points (10 years and 15 years), women with high serum urate concentrations at baseline had similar cumulative incidence of gout as men with equivalent high serum urate concentrations.

Risk of developing gout

In the Cox proportional hazard analysis, cohort, male sex, older age, non-white ethnicity and baseline serum urate were independent predictors for incident gout (table 2). Compared with baseline serum urate <6 mg/dL, the adjusted HR (95% CI) for baseline serum urate 6.0–6.9 mg/dL was 2.7 (2.0 to 3.6), for 7.0–7.9 mg/dL was 6.6 (5.0 to 8.8), for 8.0–8.9 mg/dL was 15 (11 to 20), for 8.0–8.9 mg/dL was 30 (21 to 42) and for ≥10 mg/dL was 64 (43 to 96). Sensitivity analyses did not demonstrate any major differences in the adjusted HRs when examination date or an interaction term of cohort × baseline serum urate was included in the models (online supplementary tables 5–6).

Discussion

This analysis of individual participant data demonstrates that serum urate is a strong non-linear concentration-dependent predictor of clinically evident incident gout. We provide cumulative incidence estimates that may guide discussions with individuals with hyperuricaemia about their risk of developing gout over time.

The results of this analysis including individual participant data for >18 000 people can be compared with smaller studies describing the relationship between serum urate concentrations and development of incident gout. A concentration-dependent relationship between serum urate concentrations and development of incident gout has been reported in other studies,2–5 with a wide range of estimated incidence values. Our individual participant data analysis from four separate cohorts has shown a similar incidence for men as the Normative Aging Study and a much lower annual incidence than reported in other studies.4 5 Our data indicate that the absolute risk of incident gout in those with modestly elevated serum urate concentrations is low, and even those with very high serum urate concentrations do not invariably develop gout, even over a prolonged period of follow-up. These observations imply a role for additional factors in the pathogenesis of gout; such factors would include inhibitors or promoters of crystal formation in the presence of elevated tissue urate concentrations,11 and/or genetic and environmental factors that influence the inflammatory response to deposited crystals.

We also show in the Cox proportional hazard analysis that, in addition to serum urate concentrations, sex, age and ethnicity influence the risk of developing gout. While women had lower risk of incident gout compared with men for early time points, the risk according to serum urate level became similar between men and women with longer periods of observation. The duration of exposure to elevated urate concentrations may be different in women, as serum urate concentrations are generally lower before the menopause.12 Together with the increased risk with older age, these sex findings suggest that the duration of exposure to hyperuricaemia contributes to the development of gout.

There was a small percentage of participants with baseline serum urate below 6 mg/dL who subsequently developed gout. Some of these individuals may have been misdiagnosed with gout, had a falsely low serum urate at the time of testing or had subsequent increases in serum urate. It is also possible that these individuals truly developed gout at a low serum urate concentration, potentially due to local tissue factors that influence urate solubility, crystal nucleation or crystal growth.

Limitations of this study include the variable gout definitions and methods of ascertainment. Importantly, a specific question about gout was recorded at each of the study visits included. In a previous large, multinational study, we have demonstrated that self-report of gout performs very well as a survey definition of gout, when compared with crystal identification as the gold standard.13 Our sensitivity analysis did not indicate a major difference in results depending on whether the diagnosis date or examination date was used to estimate the date of incident gout. Baseline serum urate concentrations were measured between 1972 and 1989, and it is possible that changes in the environment over the last 30 years may influence contemporary incidence estimates. In addition, the incidence estimates were based on a single urate concentration at the baseline time period, and fluctuation of serum urate over time may also influence the risk of developing disease. This study did not include other endpoints associated with hyperuricaemia, such as hypertension,14 chronic kidney disease15 or cardiovascular disease,16 and consideration of these conditions may also be of relevance when counselling people with elevated serum urate concentrations. The study addressed clinically evident gout as an outcome and was not able to examine the influence of subclinical monosodium urate (MSU) crystal deposition, which is commonly present in people with asymptomatic hyperuricaemia17 18 and is associated with more severe coronary artery disease19; future prospective studies including advanced imaging methods such as ultrasound or dual-energy CT will be instructive to address this issue.

In summary, this analysis provides cumulative incidence estimates to guide discussions with individuals with elevated serum urate concentrations about their risk of developing gout over time. Although serum urate is a strong concentration-dependent risk factor for developing incident gout, most people with hyperuricaemia do not develop clinically evident gout, even with prolonged periods of observation.

Table 2

Cox proportional hazard model analysis including cohort, age, sex, ethnicity and baseline serum urate concentration

Acknowledgments

The approval for individual-level data came from the Database of Genotype and Phenotype (DbGaP) project #834: The genetic basis of gout. The ARIC study is carried out as a collaborative study supported by the National Heart, Lung, and Blood Institute contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, N01-HC-55022, R01HL087641, R01HL59367 and R01HL086694; National Human Genome Research Institute contract U01HG004402; and National Institutes of Health contract HHSN268200625226C. Infrastructure was partly supported by Grant Number UL1RR025005, a component of the National Institutes of Health and NIH Roadmap for Medical Research. The Coronary Artery Risk Development in Young Adults Study (CARDIA) is conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with the University of Alabama at Birmingham (N01-HC95095 and N01-HC48047), University of Minnesota (N01-HC48048), Northwestern University (N01-HC48049), and Kaiser Foundation Research Institute (N01-HC48050). The FHS and the Framingham SHARe project are conducted and supported by the National Heart, Lung, and Blood Institute in collaboration with Boston University. The Framingham SHARe data used for the analyses described in this manuscript were obtained through dbGaP. The authors thank the staff and participants of the CARDIA, ARIC and FHS studies for their important contributions. This manuscript was not prepared in collaboration with, nor approved by, investigators of the CARDIA, ARIC and FHS studies and does not necessarily reflect the opinions or views of the CARDIA, ARIC and FHS studies, Boston University, or the National Heart, Lung and Blood Institute.

References

View Abstract

Footnotes

  • Handling editor Tore K Kvien

  • Contributors ND conceived and planned the study, interpreted the data, and wrote the first draft of the manuscript. ND is responsible for the overall content of the manuscript as guarantor. AP-G contributed to analysis planning, data management and manuscript drafting. CF analysed the data and contributed to analysis planning and manuscript drafting. TN, WJT and TRM contributed to study conception and planning, data interpretation, and manuscript drafting.

  • Funding This work was supported by the Health Research Council of New Zealand. TN is supported by NIH K24 AR070892.

  • Competing interests ND has received consulting fees, grants or speaker fees from Takeda, Horizon, Menarini, AstraZeneca, Ardea, Pfizer, Amgen and Kowa outside of the submitted work. WJT has received consulting fees from AstraZeneca and Pfizer outside of the submitted work. TRM has received consulting fees or grants from Horizon, AstraZeneca and Ardea outside of the submitted work. The other authors have no financial disclosures.

  • Patient consent Analysis of publicly available data.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Correction notice This article has been corrected since it published Online First. The results section of the abstract has been corrected.

  • Presented at This manuscript is based on work previously presented at the American College of Rheumatology Annual Scientific Meeting, San Francisco, USA, 2017, and was published as the following conference abstract: Dalbeth N, Phipps-Green A, Frampton C, Neogi T, Taylor WJ, Merriman TR. The Relationship between Serum Urate Concentration and Incident Gout: An Individual Participant Data Analysis [abstract]. Arthritis Rheumatol. 2017; 69 (suppl 10).

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.