You are here

  • Home
  • Archive
  • Volume 74, Issue 7
  • Lack of gene–diuretic interactions on the risk of incident gout: the Nurses’ Health Study and Health Professionals Follow-up Study

Article Text

Article menu
Clinical and epidemiological research
Concise report
Lack of gene–diuretic interactions on the risk of incident gout: the Nurses’ Health Study and Health Professionals Follow-up Study
  1. Ying Bao1,
  2. Gary Curhan1,2,
  3. Tony Merriman3,
  4. Robert Plenge2,
  5. Peter Kraft4,
  6. Hyon K Choi1,5
  1. 1Channing Division of Network Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA
  2. 2Renal Division, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA
  3. 3University of Otago, Dunedin, New Zealand
  4. 4Harvard School of Public Health, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA
  5. 5Division of Rheumatology, Allergy, and Immunology, Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA
  1. Correspondence to Dr Hyon K Choi, Division of Rheumatology, Allergy, and Immunology, Massachusetts General Hospital, Harvard Medical School, 55 Fruit Street, Bulfinch 165, Boston, MA 02114, USA; hchoi{at}partners.org

Abstract

Background Diuretic-induced gout might occur only among those with a genetic predisposition to hyperuricaemia, as suggested by a recent study with 108 self-reported gout cases.

Methods We examined the role of urate genes on the risk of diuretic-induced incident gout in 6850 women from the Nurses’ Health Study (NHS) and in 4223 men from the Health Professionals Follow-up Study (HPFS). Two published genetic risk scores (GRSs) were calculated using urate-associated single-nucleotide polymorphisms for 8 (GRS8) and 29 genes (GRS29).

Results Our analyses included 727 and 354 confirmed incident gout cases in HPFS and NHS, respectively. The multivariate relative risk (RR) for diuretic use was 2.20 and 1.69 among those with GRS8 < and ≥ the median (p for interaction=0.27). The corresponding RRs using GRS29 were 2.19 and 1.88 (p for interaction=0.40). The lack of interaction persisted in NHS (all p values >0.20) and in our analyses limited to those with hypertension in both cohorts. SLC22A11 (OAT4) showed a significant interaction only among women but in the opposite direction to the recent study.

Conclusions In these large prospective studies, individuals with a genetic predisposition for hyperuricaemia are not at a higher risk of developing diuretic-induced gout than those without.

  • Epidemiology
  • Gout
  • Gene Polymorphism

https://doi.org/10.1136/annrheumdis-2014-206534

Statistics from Altmetric.com

    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.

    Introduction

    Gout is a common and excruciatingly painful inflammatory arthritis caused by hyperuricaemia. In addition to various environmental risk factors for gout, substantial genetic predisposition has also been recognised.1 The majority of the eight genes discovered by the initial genome-wide association studies (GWAS) are involved in the renal urate-transport system.1–3 In 2013, the Global Urate Genetics Consortium identified and replicated 20 additional genome-wide significant loci in association with serum urate concentrations.4

    Diuretics, particularly thiazide and loop diuretics, increase the risk of incident gout,5 ,6 likely through urate transporters (eg, OAT4)3 and volume depletion promoting urate reabsorption.7 If diuretic use differentially affects the risk of gout according to a certain genetic predisposition for hyperuricaemia, such genes could be used to predict gout risk in relation to diuretic use. Indeed, a recent analysis based on the Atherosclerosis Risk in Communities (ARIC) study (with 108 self-reported incident cases of gout) reported that diuretic-induced gout occurs only among those with a genetic predisposition to hyperuricaemia.3 To our knowledge, no study has replicated or refuted these findings. To examine these issues with a large number of confirmed cases of incident gout (N=1081), we analysed GWAS data obtained from participants in the Nurses’ Health Study (NHS) and the Health Professionals Follow-up Study (HPFS).

    Participants and methods

    Study population

    As described previously in detail,8 ,9 NHS and HPFS are prospective cohorts of women and men, respectively, who have been administered the validated food frequency questionnaire (FFQ) every four years, as well as biennial questionnaires for new disease diagnoses and drug use (including diuretics). The current analysis included 6850 and 4223 initially healthy women and men, respectively, of European ancestry for whom genotype data based on GWAS were available10 and who were gout-free at baseline.

    Assessment of incident gout cases, diuretic use and covariates

    We ascertained incident gout cases using the American College of Rheumatology (ACR) gout survey criteria, as described previously.8 ,9 In NHS, diuretic use (ie, thiazides and loop diuretics) was determined in 1980, 1982, 1988, 1994 and every two years thereafter. In HPFS, diuretic use was determined in 1986 and every two years thereafter. Baseline and biennial follow-up questionnaires inquired about physician-diagnosed hypertension. A participant was considered to have hypertension from the time first reported on the questionnaire. Dietary covariates were obtained via validated FFQs. Information on non-dietary factors (eg, medication or supplement use, medical conditions, menopausal status and postmenopausal hormone use) was assessed every two to four years.6 ,8 ,9 ,11–15

    Genotyping and genetic risk score

    We selected single-nucleotide polymorphisms (SNPs) known to be associated with serum urate levels. Two sets of genetic risk scores (GRSs) were calculated as published previously; one based on 8 SNPs2 (GRS8, used by the ARIC study above3) and the other based on 29 SNPs (GRS29, a new score incorporating additional novel genes4) (see online supplementary table S1). Higher scores indicate a greater genetic predisposition to higher serum uric acid levels. Of note, the aforementioned ARIC study analysis used GRS8,3 which mostly consisted of urate transporters.1–3 SNP genotyping and imputation have been described in detail elsewhere.4 ,16–20 Most of the SNPs were genotyped or had a high-imputation quality score (MACH r2≥0.8).10

    Statistical analysis

    We examined the association between diuretic use and the risk of gout, according to the median of GRS, using Cox proportional hazards models adjusting for previously identified risk factors.6 ,8 ,9 ,11–15 Information on use on diuretic use was updated according to the biennial questionnaires as a time-varying exposure. Tests for interaction by GRS were performed by the Wald test of a cross-product term of GRS (≥ median vs < median) and diuretic use (yes or no). We then repeated the analyses among participants with hypertension. All statistical analyses were performed with SAS V.9.1. All p values are two sided.

    Results

    At baseline, the distribution of gout risk factors was similar between the group with GRS < median and the group with GRS ≥ median (table 1). Further, at baseline, those using diuretics were older and more likely to have a higher body mass index (BMI) and hypertension; other characteristics were similar (see online supplementary table S2). GRS83 ranged from −3.67 to 4.24, and GRS294 ranged from 12.2 to 44.5. In both cohorts, participants with a higher GRS had a higher risk of incident gout (see online supplementary table S3).

    View this table:
    Table 1

    Baseline gout risk factors by genetic scores in NHS and HPFS participants

    In HPFS, compared with no use, the multivariate relative risk (RR) of gout for thiazide or loop diuretic use was 2.20 (95% CI 1.42 to 3.42) among those with GRS8 < median and 1.69 (95% CI 1.21to 2.37) among those with GRS8 ≥ median (p for interaction=0.27) (table 2). The corresponding multivariate RRs according to GRS29 median value categories were 2.19 (95% CI 1.40 to 3.43) among those with GRS29 < median and 1.88 (95% CI 1.34 to 2.63) among those with GRS29 ≥ median (p for interaction=0.40). Similar results were found in NHS (table 2).

    View this table:
    Table 2

    RR of gout by diuretic use, stratified by median genetic scores

    We additionally evaluated two individual urate genes previously shown to have a significant interaction with diuretic use (SLC2A9 and SLC22A11).3 A significant interaction was observed only for SLC22A11 among women but in the opposite direction to the ARIC study finding3 (table 3).

    View this table:
    Table 3

    RR of gout by diuretic use, stratified by SLC2A9 (rs13129697) and SLC22A11 (rs2078267)

    Further analyses restricted to participants with hypertension yielded similar results to the main analyses in both cohorts (see online supplementary tables S4–8).

    Discussion

    In these two large prospective studies of white adults, the risk of gout induced by thiazide or loop diuretics did not vary according to the genetic risk for elevated serum urate. This lack of significant interaction was evident using the previously employed genetic score based on eight SNPs,3 as well as the score that included the most updated list of SNPs.4 Similarly, no significant interaction existed with the individual genes previously shown to have significant interactions, except for SLC22A11 among women, which showed an interaction opposite to the ARIC study finding.3 These findings persisted in the subgroups of those with hypertension. Collectively, our findings reject the hypothesis that individuals with a genetic predisposition for hyperuricaemia are at a higher risk of developing diuretic-induced gout compared with those without such a predisposition.

    Our findings do not replicate the aforementioned report based on the ARIC study cohort.3 Several aspects of that study notably differ from the current study. The ARIC analysis analysed a relatively small number of cases for a subgroup/interaction analysis (ie, 108 patients with gout and hypertension),3 whereas our cohorts had substantially more gout cases (1081 cases, including 506 with hypertension). The ARIC study had only one gout case in the group of thiazide use and four cases in the group of thiazide or loop diuretic use among those with GRS8 < median. These small sample sizes likely contributed to the main conclusion of the paper: a null or inverse impact of thiazide or loop diuretic use among those with GRS < median. In our study, each group used for comparison had substantially more cases than the ARIC study, including the subgroup analyses limited to those with hypertension. Furthermore, our gout cases were confirmed based on the ACR survey criteria,8 ,9 whereas the ARIC study was based on self-reported diagnosis.3

    Not being able to replicate the findings by our two larger cohorts underscores the need for an appropriate sample size in this type of interaction analysis, as well as the value of replication data (preferably in the same paper, as is now commonly required in papers reporting on genetics). This appears particularly relevant as many published subgroup/interaction genetic analyses have not been replicated in follow-up studies. Indeed, we found that the significant interaction with SLC22A11 among women was in the opposite direction to the ARIC study finding, which is an extreme example of this issue. This inconsistency may reflect a difference in the sample sizes, multiple testing or random chance, and would not justify biological speculations.

    Several strengths and potential limitations of our study deserve comment. The strengths include the use of large cohort studies with long-term follow-up, multiple measures of diuretic use and BMI, comprehensive measures of diet and lifestyle factors, and GRS combining genetic information from the most updated list of variants associated with serum urate levels.4 Furthermore, the overall consistent findings of the lack of interaction using GRSs in our two cohorts indicate the robustness of our results. Certain measurement errors in dietary factors (our covariates) are inevitable, but FFQs have been well validated in our cohorts and have been successfully used for many dietary risk factor studies of gout. Furthermore, the participants in our study were middle-aged and older adults of European ancestry recruited in the USA, and it is unknown whether our findings could be generalised to other demographic or ethnic groups.

    In conclusion, these prospective studies based on a large number of incident cases of confirmed gout suggest that individuals with a genetic predisposition for hyperuricaemia are not at a greater risk of developing diuretic-induced gout than those without such a predisposition. These findings do not appear to support the potential utility of these genes to assess gout risk in relation to diuretic use.

    References

    1. ↵
      1. Reginato AM,
      2. Mount DB,
      3. Yang I, et al
      . The genetics of hyperuricaemia and gout. Nat Rev Rheumatol 2012;8:610–21. doi:10.1038/nrrheum.2012.144
      OpenUrlCrossRefPubMed
    2. ↵
      1. Yang Q,
      2. Kottgen A,
      3. Dehghan A, et al
      . Multiple genetic loci influence serum urate levels and their relationship with gout and cardiovascular disease risk factors. Circ Cardiovasc Genet 2010;3:523–30. doi:10.1161/CIRCGENETICS.109.934455
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. McAdams-DeMarco MA,
      2. Maynard JW,
      3. Baer AN, et al
      . A urate gene-by-diuretic interaction and gout risk in participants with hypertension: results from the ARIC study. Ann Rheum Dis 2013;72:701–6. doi:10.1136/annrheumdis-2011-201186
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Kottgen A,
      2. Albrecht E,
      3. Teumer A, et al
      . Genome-wide association analyses identify 18 new loci associated with serum urate concentrations. Nat Genet 2013;45:145–54. doi:10.1038/ng.2500
      OpenUrlCrossRefPubMed
    5. ↵
      1. Choi HK,
      2. Soriano LC,
      3. Zhang Y, et al
      . Antihypertensive drugs and risk of incident gout among patients with hypertension: population based case-control study. BMJ 2012;344:d8190. doi:10.1136/bmj.d8190
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Choi HK,
      2. Atkinson K,
      3. Karlson EW, et al
      . Obesity, weight change, hypertension, diuretic use, and risk of gout in men: the health professionals follow-up study. Arch Intern Med 2005;165:742–8. doi:10.1001/archinte.165.7.742
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Steele TH,
      2. Oppenheimer S
      . Factors affecting urate excretion following diuretic administration in man. Am J Med 1969;47:564–74. doi:10.1016/0002-9343(69)90187-9
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Choi HK,
      2. Atkinson K,
      3. Karlson EW, et al
      . Purine-Rich Foods, Dairy and Protein Intake, and the Risk of Gout in Men. New Eng J Med 2004;350:1093–103. doi:10.1056/NEJMoa035700
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Choi HK,
      2. Willett W,
      3. Curhan G
      . Fructose-rich beverages and risk of gout in women. JAMA 2010;304:2270–8. doi:10.1001/jama.2010.1638
      OpenUrlCrossRefPubMed
    10. ↵
      1. Qi Q,
      2. Chu AY,
      3. Kang JH, et al
      . Fried food consumption, genetic risk, and body mass index: gene-diet interaction analysis in three US cohort studies. BMJ 2014;348:g1610. doi:10.1136/bmj.g1610
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Choi HK,
      2. Atkinson K,
      3. Karlson EW, et al
      . Alcohol Intake and Risk of Incident Gout in Men—A Prospective Study. Lancet 2004;363:1277–81. doi:10.1016/S0140-6736(04)16000-5
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Choi HK,
      2. Curhan G
      . Soft drinks, fructose consumption, and the risk of gout in men: prospective cohort study. BMJ 2008;336:309–12. doi:10.1136/bmj.39449.819271.BE
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Choi HK,
      2. Curhan G
      . Coffee consumption and risk of incident gout in women: the Nurses’ Health Study. Am J Clin Nutr 2010;92:922–7. doi:10.3945/ajcn.2010.29565
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Choi HK,
      2. Gao X,
      3. Curhan G
      . Vitamin C intake and the risk of gout in men: a prospective study. Arch Intern Med 2009;169:502–7. doi:10.1001/archinternmed.2008.606
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Choi HK,
      2. Willett W,
      3. Curhan G
      . Coffee consumption and risk of incident gout in men: A prospective study. Arthritis Rheum 2007;56:2049–55. doi:10.1002/art.22712
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Cornelis MC,
      2. Monda KL,
      3. Yu K, et al
      . Genome-wide meta-analysis identifies regions on 7p21 (AHR) and 15q24 (CYP1A2) as determinants of habitual caffeine consumption. PLoS Genet 2011;7:e1002033. doi:10.1371/journal.pgen.1002033
      OpenUrlCrossRefPubMed
    17. ↵
      1. Jensen MK,
      2. Pers TH,
      3. Dworzynski P, et al
      . Protein interaction-based genome-wide analysis of incident coronary heart disease. Circ Cardiovasc Genet 2011;4:549–56. doi:10.1161/CIRCGENETICS.111.960393
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Wiggs JL,
      2. Kang JH,
      3. Yaspan BL, et al
      . Common variants near CAV1 and CAV2 are associated with primary open-angle glaucoma in Caucasians from the USA. Hum Mol Genet 2011;20:4707–13. doi:10.1093/hmg/ddr382
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Hunter DJ,
      2. Kraft P,
      3. Jacobs KB, et al
      . A genome-wide association study identifies alleles in FGFR2 associated with risk of sporadic postmenopausal breast cancer. Nat Genet 2007;39:870–4. doi:10.1038/ng2075
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Qi L,
      2. Cornelis MC,
      3. Kraft P, et al
      . Genetic variants at 2q24 are associated with susceptibility to type 2 diabetes. Hum Mol Genet 2010;19:2706–15. doi:10.1093/hmg/ddq156
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Handling editor Tore K Kvien

    • Contributors YB and HKC had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the analysis. YB, GC, RP, PK and HKC were responsible for the overall study design and data analysis. All authors were involved in interpretation, preparation and critical review of the manuscript.

    • Funding This research was supported by grants R01AR056291, R01AR065944, UM1CA167552 and P01CA87969 from the National Institutes of Health.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Partners HealthCare System institutional review board.

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