Abstract
Background
Patients with hyperuricemia or gout often have metabolic syndrome. Few prospective studies examined the risk of incident diabetes mellitus (DM) in patients with gout, and no data exist whether the DM risk in gout differs by sex.
Methods
Using data from a US commercial insurance plan (2003–2012), we conducted a cohort study to examine the overall and sex-specific incidence rate (IR) of DM in patients aged ≥40 years with gout compared to those with osteoarthritis. Incident DM was defined based on a diagnosis of DM and a dispensing for anti-diabetic drugs. We tested the sex-specific effect of gout on DM risk.
Results
The study cohort consisted of 54,075 gout and 162,225 osteoarthritis patients, matched on age, sex and index date. The mean age was 56.2 years and 84.8% were men. Over a mean follow-up of 1.9 years, the IR of DM was 1.91 per 100 person-years in gout and 1.12 per 100 person-years in osteoarthritis patients. After adjusting for age, comorbidities, medications, and health care utilization, gout was associated with an increased risk of DM (hazard ratio [HR] 1.45, 95%CI 1.37–1.54) for both sexes. The impact of gout on the risk of incident DM was greater in women (HR 1.78, 95%CI 1.51–2.09) than men (HR 1.41, 95%CI 1.33–1.50) with a significant interaction between sex and gout (p=0.0009).
Conclusion
Gout was associated with an increased risk of developing DM compared with osteoarthritis after adjusting for potential confounders, and the risk associated with gout was higher among women than men.
Keywords: gout, diabetes mellitus, uric acid, sex
INTRODUCTION
Gout is a common inflammatory arthritis caused by hyperuricemia.[1] In 2007, nearly 8.3 million Americans had gout and the prevalence of gout and hyperuricemia continues to increase.[2] Cross-sectional studies report that gout and hyperuricemia are associated with metabolic syndrome generally defined as a cluster of cardiovascular risk factors including abdominal obesity, hypertension, dyslipidemia, and insulin resistance.[3, 4] A recent Taiwanese study reported that gout and type 2 DM shared common genetic risk alleles.[5] While growing evidence supports that hyperuricemia is independently associated with a greater future risk of metabolic syndrome and diabetes mellitus (DM),[6–9] limited longitudinal data exist on the risk of incident DM in patients with gout.[10] One prospective cohort of men with a high cardiovascular risk profile found that gout associated with a higher risk of incident DM compared to those with no gout; no analyses were performed on women.[10]
Since gout and hyperuricemia are more common in men than women, most epidemiologic studies focus on men. But, several previous studies observed a greater risk of coronary heart disease and hypertension associated with hyperuricemia or gout in women compared to men.[11–13] Although it is not well-understood, the different impact of gout or hyperuricemia on cardiovascular risk by sex is thought to be related to differences in serum uric acid levels and distribution of comorbidities between sexes.[11, 12, 14, 15] In addition, uric acid was a strong independent risk factor of DM in women but not in men in a previous population-based cohort study.[16] To date, no prospective data are available on whether the risk of incident DM or metabolic syndrome associated with gout is different between women and men.
We therefore studied a large longitudinal cohort from a US health care utilization database with the following aims: 1) to estimate the overall and sex-specific incidence rate (IR) of DM among patients with and without gout, 2) to assess the risk of incident DM among gout compared with non-gout patients, and 3) to test whether the risk of incident DM associated with gout differs by sex.
METHODS
Data Source
We conducted a cohort study using the claims data from United HealthCare, a large commercial U.S. health plan, for the period January 1, 2003 to December 31, 2012. This database contains longitudinal claims information including medical diagnoses, procedures, hospitalizations, physician visits, and pharmacy dispensing on more than 13 million fully-insured subscribers with medical and pharmacy coverage at any particular time point across the United States. Results for outpatient laboratory tests including hemoglobin A1c (HbA1c) and uric acid levels were available on a subset of beneficiaries. Personal identifiers were removed from the dataset before the analysis to protect subject confidentiality. Patient informed consent was therefore not required. The study protocol was approved by the Institutional Review Board of Brigham and Women’s Hospital.
Study Cohort
Eligible patients for the gout group consisted of subjects aged 40 years and older who had at least 2 visits coded with the International Classification of Diseases, 9th Revision, Clinical Modification (ICD 9-CM) code, 274.0X, 274.8X and 274.9 for gout. The index date for the ‘gout group’ was the date of the first dispensing of a gout-related drug (i.e., allopurinol, febuxostat, colchicine, probenecid, selective and non-selective non-steroidal anti-inflammatory drugs (NSAIDs), and systemic and intra-articular steroids) after at least 365 days of continuous health plan eligibility; thus, all persons in the gout cohort were required to have had two diagnoses and filled at least one prescription for gout at the index date. We excluded nursing home residents or patients who had a malignancy diagnosis or received chemotherapy in the 365-day period prior to the index date. To ensure that only incident cases of DM were included, we excluded patients with a diagnosis of DM, dispensings for anti-diabetic drugs or use of insulin pump in the 365-day period prior to the index date from both cohorts.
We selected two non-gout groups aged 40 years and older for comparison. The first comparison group, described as the ‘osteoarthritis’ group, consisted of patients who had at least 2 visits coded with the ICD-9 code, 715.XX for osteoarthritis, a chronic arthritis with less systemic inflammation. This group was chosen as the primary comparison group, because patients with osteoarthritis have health care utilization (e.g. frequency of visits to physicians) and baseline characteristics such as physical activity and obesity similar to the gout group. The index date for the osteoarthritis group began at the first receipt of opioids, selective or non-selective NSAIDs after at least 2 physician visits for osteoarthritis. The aforementioned exclusion criteria were then applied. Patients with osteoarthritis were not allowed to have a diagnosis of gout or use of allopurinol, febuxostat, colchicine, or probenecid in the 365-day period prior to or on the index date. The osteoarthritis group was then matched to the gout group on age, sex and index date with a 3:1 ratio.
The 2nd comparison group, described herein as the ‘non-gout group’, consisted of patients who did not have a diagnosis gout at baseline and had at least 2 physician visits after at least 12 months of continuous health plan eligibility. The index date for the non-gout patients began at the first receipt of any prescription drugs after at least 2 physician visits. Non-gout patients were not allowed to use allopurinol, febuxostat, colchicine, or probenecid in the 365-day period prior to or on the index date. The non-gout group was then matched to the gout group on age, sex and index date with a 3:1 ratio.
Patients in both groups were followed from the index date to the first of any of the following censoring events: development of DM, development of gout (for both osteoarthritis and non-gout groups), insurance disenrollment, end of study period, or death.
Outcome Definition
Incident DM was defined based on a diagnosis of DM and a dispensing for an anti-diabetic drug (Appendix 1). The date of the first anti-diabetic drug dispensing was used as the date of outcome occurrence.
Covariates
Patients’ baseline variables potentially related to gout or development of DM were examined using data from the 365 days before the index date. These variables included demographic factors (age, sex and region of residence), comorbidities (hypertension, stroke, coronary heart disease, heart failure, chronic kidney disease, liver disease, sleep apnea, hyperlipidemia, alcoholism, smoking, and obesity), medications (steroids, selective and non-selective NSAIDs, angiotensin converting enzyme inhibitors, angiotensin II receptor blockers, and diuretics, lipid-lowering drugs), markers of health care utilization intensity (number of visits to any physicians, acute care hospitalization, emergency room visits, and number of different prescription drugs) and ordered outpatient laboratory tests such as renal function test, HbA1c, acute phase reactants, and uric acid. To quantify patients’ comorbidities, we also calculated a comorbidity score that combined conditions included in both the Charlson Index and the Elixhauser system based on ICD-9-CM.[17] The comorbidity score is a summative score, based on 20 major medical conditions such as metastatic cancer, congestive heart failure, dementia, renal failure, weight loss, hemiplegia, pulmonary and liver disease and ranges from −2 to 26. In a subgroup of the study cohort, outpatient laboratory test results such as HbA1c, acute phase reactants, and uric acid levels were also examined in a subgroup of the study cohort.
Statistical Analyses
We compared the baseline characteristics between the gout and osteoarthritis groups matched on age, sex, and index date. We estimated the overall and sex-specific incidence rates (IR) of DM with 95% confidence interval (CI) in both groups. The overall and sex-specific rate ratios (RR) with 95% CI were estimated by dividing the incidence rates of DM among gout patients by those of osteoarthritis.[18] We tested for an interaction between gout and sex by including the product term in a multivariable Cox proportional hazards model and found a statistically significant interaction (p=0.0009). We therefore used sex-stratified multivariable Cox models adjusting for various potential confounders to compare the risk of incident DM among gout patients to those with osteoarthritis.[19] Partially adjusted, sex-stratified Cox models included age, comorbidity score and number of prescription drugs. Fully adjusted Cox models included age, comorbidities, medications, and health care utilization factors listed in Table 1. All the analyses were then repeated for the comparison between the gout and non-gout groups.
Table 1.
Baseline characteristics of study cohorts in the 365 days before study entry.
| Gout | Osteoarthritis | |
|---|---|---|
| N | 54,075 | 162,225 |
| Percentage or mean ± standard deviation | ||
| Follow-up, year | 1.9 ± 1.7 | 1.9 ± 1.7 |
| Demographic * | ||
| Age, years | 56.2 ± 9.2 | 56.2 ± 9.2 |
| Men | 84.8 | 84.8 |
| Comorbidities | ||
| Hypertension | 66.7 | 48.0 |
| Stroke | 4.3 | 4.1 |
| Coronary heart disease | 12.5 | 11.2 |
| Heart failure | 4.1 | 1.9 |
| Lung disease | 11.1 | 12.9 |
| Chronic kidney disease | 8.2 | 1.8 |
| Liver disease | 3.7 | 2.8 |
| Hyperlipidemia | 59.1 | 50.0 |
| Obesity | 9.0 | 8.3 |
| Sleep apnea | 5.1 | 5.3 |
| Smoking | 6.3 | 10.5 |
| Alcoholism | 1.0 | 0.8 |
| Comorbidity score a | 0.1 ± 1.4 | 0.1 ± 1.0 |
| Medications | ||
| Lipid-lowering drugs | 39.0 | 33.5 |
| Recent steroid use | 7.1 | 4.1 |
| Cumulative prednisone-equivalent dose (mg) | 117.2 ± 431.8 | 80.2 ± 552.5 |
| NSAIDs | 57.1 | 53.3 |
| Selective cox-2 inhibitors | 6.2 | 12.8 |
| ACE inhibitors or ARBs | 41.9 | 26.8 |
| Diuretics | 22.5 | 13.0 |
| Health care utilization | ||
| No. of outpatient physician visits | 6.9 ± 5.4 | 8.0 ± 5.7 |
| Emergency room visit | 21.6 | 19.5 |
| Acute hospitalization | 13.6 | 21.4 |
| No. of prescription drugs | 8.2 ± 5.0 | 7.7 ± 4.9 |
| Laboratory test | ||
| BUN ordered | 54.6 | 43.2 |
| Creatinine ordered | 56.1 | 44.2 |
| C-reactive protein ordered | 9.7 | 7.8 |
| ESR ordered | 16.6 | 11.5 |
| HbA1c ordered | 9.8 | 7.0 |
| HbA1c level available | 3.4 | 2.4 |
| HbA1c, % b (median, IQR) | 5.8 ± 0.5 (5.8, 5.5–6.0) | 5.8 ± 0.7 (5.7, 5.5–6.0) |
| Uric acid ordered | 56.4 | 6.7 |
| Uric acid level available | 20.2 | 2.3 |
| Uric acid, mg/dl c (median, IQR) | 7.9 ± 2.0 (8.0, 6.5–9.2) | 5.9 ± 1.4 (5.8, 4.9–6.9) |
ACE: angiotensin-converting enzyme, ARB: angiotensin receptor blocker, NSAID: nonsteroidal anti-inflammatory drug, BUN: blood urea nitrogen, ESR: erythrocyte sedimentation rate, HbA1c: hemoglobin A1c, IQR: interquartile range
Gout and osteoarthritis groups are age-, sex- and index date-matched.
The range of the comorbidity score is −2 to 26.
The mean HbA1c level was calculated among patients with baseline HbA1c levels available.
The mean uric acid level was calculated among patients with baseline uric acid levels available.
Kaplan-Meier curves were plotted by sex for the cumulative incidence of DM in the gout and matched osteoarthritis group. A sensitivity analysis adjusted for baseline HbA1c was conducted among patients in whom we had a baseline HbA1c level available. In addition, a sensitivity analysis adjusted for elevated acute phase reactants at baseline was conducted among patients in whom we had a baseline acute phase reactant level available. We also used stratified Cox proportional hazards regression models to assess whether impact of gout on the risk of DM differs by steroid use at baseline, and cumulative steroid dose during the 365-day baseline period.
The proportional hazards assumption was assessed by testing the significance of the interaction term between exposure (i.e., gout or osteoarthritis) and time, and was not violated in any models.[20] All analyses were done using SAS 9.3 Statistical Software (SAS Institute Inc., Cary, NC).
RESULTS
Cohort Selection
There were initially 289,684 patients with at least one gout diagnosis and 1,822,117 with at least one osteoarthritis diagnosis after a 365-day enrollment period. After applying the inclusion and exclusion criteria, we selected 3 osteoarthritis patients matched to each gout patient on age, sex, and index date. Our final study cohort includes 54,075 gout and 162,225 osteoarthritis patients. Figure 1 displays the study cohort selection process. The mean (SD) follow-up time was 1.94 (1.72) years for gout and 1.91 (1.72) years for osteoarthritis patients.
Figure 1. Cohort selection flow.
The final study cohort included 54,075 patients with gout and 162,225 with osteoarthritis matched on age, sex and index date.
For the comparison between gout and non-gout groups, 66,119 gout patients matched on age, sex, and index date to non-gout patients with a 1:3 ratio were included (Appendix 2). The mean (SD) follow-up time was 1.86 (1.68) years for gout and 1.87 (1.74) years for non-gout patients.
Patient Characteristics
Baseline characteristics of the age-, sex- and index date-matched groups were compared (Table 1). The mean age was 56.2 years and men comprised 84.8% in both groups. Patients with gout experienced a greater frequency of hypertension, heart failure, coronary heart disease, chronic kidney disease, obesity, and hyperlipidemia compared to the osteoarthritis group. Use of drugs such as steroids, NSAIDs and diuretics and emergency room visits were more common in gout than osteoarthritis. During the 365-day baseline period, HbA1c tests were ordered in 9.8% of gout patients and 7.0% of osteoarthritis. Among patients with gout, women had more comorbidities such as hypertension, obesity, chronic kidney disease and heart failure, and used steroids and health care system more frequently than men did (Appendix 3). Sixty percent of gout patients received at least one prescription for either allopurinol or febuxostat at baseline.
Risk of Incident DM in gout vs. osteoarthritis
The IR of DM was 1.91 per 100 person-years in gout and 1.12 per 100 person-years in osteoarthritis patients (Table 2). The age-, sex- and index date-matched rate ratio (RR) was 1.71 (95% CI 1.62–1.81) in the overall gout group compared to the osteoarthritis, but, importantly the RR was higher in women (2.42, 95% CI 2.09–2.81) than men (1.62, 95% CI 1.53–1.72). Table 3 summarizes the results from various multivariable Cox regression analyses comparing the gout to the osteoarthritis group. After adjusting for age, comorbidities, medications, and health care utilization intensity, the DM risk associated with gout was increased in both men and women, with a greater HR in women (1.78, 95% CI 1.51–2.09) compared to men (1.41, 95% CI 1.33–1.50). Appendix 4 shows Kaplan-Meier curves comparing the cumulative incidence of DM in the age-, sex- and index date-matched gout and osteoarthritis groups.
Table 2.
Incidence rates of diabetes in patients with gout compared to those with osteoarthritis and non-gout.
| Gout | Osteoarthritis | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| No. of patients | Diabetes cases | Person-years | Incidence rate (95% CI) | No. of patients | Diabetes cases | Person-years | Incidence rate (95% CI) | Rate ratio* (95% CI) | |
| All | 54,075 | 2,000 | 104,772 | 1.91 (1.83–2.00) | 162,225 | 3,466 | 309,795 | 1.12 (1.08–1.16) | 1.71 (1.62–1.81) |
| Women | 8,224 | 305 | 13,782 | 2.21 (1.98–2.48) | 24,672 | 400 | 43,674 | 1.15 (1.11–1.19) | 2.42 (2.09–2.81) |
| Men | 45,851 | 1,695 | 90,991 | 1.86 (1.78–1.95) | 137,553 | 3,066 | 266,121 | 0.92 (0.83–1.01) | 1.62 (1.53–1.72) |
CI: confidence interval, Incidence rate is per 100 person-years.
Gout and osteoarthritis groups are age-, sex- and index date-matched.
Table 3.
Risk of incident diabetes in patients with gout compared to those with osteoarthritis.
| Adjustment | Hazard ratio (95% CI) | |
|---|---|---|
| Women | Age | 2.42 (2.09–2.81) |
| Age, comorbidity score and number of prescription drugs | 2.32 (1.99–2.69) | |
| Fully adjusted* | 1.78 (1.51–2.09) | |
| Men | Age | 1.61 (1.52–1.71) |
| Age, comorbidity score and number of prescription drugs | 1.59 (1.49–1.68) | |
| Fully adjusted* | 1.41 (1.33–1.50) |
CI: confidence interval
The full model includes age, comorbidities, medications, and health care utilization factors listed in Table 1.
Risk of Incident DM in gout vs. non-gout
The IR of DM was 1.83 per 100 person-years in gout and 0.98 per 100 person-years in non-gout patients. The age-, sex- and index date-matched rate ratio (RR) was 1.88 (95% CI 1.78–1.98) in the gout group compared to the non-gout. Similar to the comparison between gout and osteoarthritis, the RR was higher in women (2.74, 95% CI 2.35–3.19) than men (1.79, 95% CI 1.69–1.89). After adjusting for age, comorbidities, medications, and health care utilization intensity, the DM risk associated with gout was increased in both men and women, with a greater HR in women (1.58, 95% CI 1.29–1.94) compared to men (1.27, 95% CI 1.18–1.36).
Sensitivity Analysis
Among the subgroup of patients with baseline HbA1c levels available, gout compared to OA appears to be associated with an increased risk of incident DM in both women (HR 1.71, 95% CI, 0.85–3.43) and men (HR 1.37, 95% CI 1.06–1.76) after adjusting for age and baseline HbA1c. Among the subgroup of patients with baseline acute phase reactant levels available, gout was still associated with an increased risk of incident DM in both women (HR 2.66, 95% CI, 1.44–4.92) and men (HR 1.54, 95% CI 1.19–1.98) compared to OA patients after adjusting for age and baseline acute phase reactants. The risk of incident DM associated with gout was consistently increased in both women and men, regardless of baseline steroid use and cumulative prednisone-equivalent dose (Table 4). The risk of incident DM associated with gout increased in women who had a higher cumulative prednisone-equivalent dose during the 365-day baseline period, with the HR of 2.29 (95% CI 1.40–3.74) in the highest tertile (cumulative dose ≥325mg). In men with the highest cumulative prednisone-equivalent dose (≥316mg), the HR of incident DM associated with gout was 1.27 (95% CI 1.01–1.60).
Table 4.
Sensitivity analysis: Risk of incident diabetes in patients with gout compared to those with osteoarthritis by baseline steroid use.
| Steroid use at baseline | Hazard ratio (95% CI)* | |
|---|---|---|
| Women | No | 1.79 (1.47–2.19) |
| Yes | 1.70 (1.26–2.29) | |
| Cumulative prednisone-equivalent dose a | ||
| Tertile 1: 5 to 105 mg | 1.17 (0.68–2.01) | |
| Tertile 2: 107 to 320 mg | 1.86 (1.03–3.37) | |
| Tertile 3: 325 mg or more | 2.29 (1.40–3.74) | |
|
| ||
| Men | No | 1.42 (1.32–1.53) |
| Yes | 1.38 (1.20–1.57) | |
| Cumulative prednisone-equivalent dose a | ||
| Tertile 1: 5 to 105 mg | 1.41 (1.12–1.77) | |
| Tertile 2: 106 to 315 mg | 1.52 (1.19–1.95) | |
| Tertile 3: 316 mg or more | 1.27 (1.01–1.60) | |
CI: confidence interval
The full model includes age, comorbidity index, comorbidities, medications, and health care utilization factors listed in Table 1.
Cumulative prednisone-equivalent dose was calculated based on the total dose of steroids during the 365-day baseline period.
DISCUSSION
In this large cohort study using data from a nationwide US commercial insurer, nearly 2% of the patients with gout developed type 2 DM during a mean follow-up of 1.9 years. Gout was associated with an increased risk of developing DM after controlling for potential confounders including age, comorbidities, medications, and health care use patterns. This result is consistent with a previous study based on male patients in the Multiple risk Factor Intervention Trial.[10] Furthermore, the results of this present study draw attention to a greater impact of gout on the risk of incident DM in women compared to men. This study also showed that women with gout had more comorbid conditions and used prescription drugs and health care system more often than men. Our results were robust in the analyses from the comparison between gout and non-gout as well as sensitivity analyses adjusting for baseline HbA1c, and stratified by the baseline use of steroids.
The reason why women with gout have a higher risk of DM than men with gout is unclear. A prior study showed a greater impact of hyperuricemia on the risk of metabolic syndrome in women versus men.[8] Similarly, a number of previous studies observed sex-specific effects of gout on cardiovascular risk.[11–13] In a Canadian population-based study, women with gout had a relative risk of 1.39 for acute myocardial infarction while men with gout had a relative risk of 1.11, compared to those without gout.[11] Women with gout were noted to have a higher mean serum uric acid concentration and a lower mean urinary excretion of uric acid than did men with gout in a Spanish cohort of men and postmenopausal women.[14] It has been suggested that these differences in uric acid levels and urate metabolism between sexes might explain more pronounced cardiovascular risks associated with gout in women than men.[14, 15] In the present study, substantial differences exist in the baseline characteristics of men and women with gout. The mean age was higher in women than men, and the proportion of patients with obesity, smoking and recent steroid use was much higher in women with gout compared to men with gout. The impact of gout on the risk of DM remained larger in women even after adjusting for these differences in baseline characteristics. However, it is possible that the degree of confounding by unmeasured variables such as physical activity, diet pattern, and body mass index, on the association between gout and DM is larger in women than men.
Several prospective studies showed that hyperuricemia preceded the development of hyperinsulinemia, suggesting hyperuricemia as an independent risk factor for hyperinsulinemia or DM.[21, 22] Furthermore, markers of systemic inflammation such as C-reactive protein, interleukin-6 and soluble adhesion molecules appear to play a role in the development of type 2 DM.[23–26] Thus, systemic inflammation caused by gout may contribute to insulin resistance at the onset of DM. Another hypothesis generated by the present study posits a potential role for urate-lowering therapy in reducing the risk of incident DM among patients with chronic gout. No published studies present evidence on the effect of urate-lowering therapy on the risk of DM. However, a small cohort study of 12 end-stage renal disease patients showed a significant decrease in serum low-density lipoprotein cholesterol and triglycerides after a treatment with allopurinol 300mg twice daily for 3 months.[27] Furthermore, a potential benefit of allopurinol in hypertension, cardiovascular morbidity and mortality has been suggested. [28, 29]
Several strengths of this study are noteworthy. First, we examined a large cohort of gout and osteoarthritis patients in a population that is representative of the U.S. commercially-insured population. Second, to minimize surveillance bias, we selected a group of patients with osteoarthritis, a chronic non-inflammatory medical condition requiring pharmacologic treatment. In addition, we had an additional comparison between gout and non gout and found similarly increased risk of incident DM associated with gout. Third, we included a large number of women in the cohort and found an interaction between sex and gout on incident DM. Fourth, unlike most claims data sources, we had actual laboratory results (e.g., HbA1c and uric acid) available in a subgroup of patients. Sensitivity analyses in this subgroup of patients with baseline HbA1c levels available showed consistently increased risk of incident DM in patients with gout. Lastly, we also performed additional analyses to examine the effect of gout stratified by the use and cumulative dose of systemic steroids.
There are limitations to this study. First, we mainly relied on the claims data to select patients with gout, osteoarthritis or non-gout and identify their comorbidities, raising the possibility for exposure and outcome misclassification bias. However, we required all the patients in the study cohort to have at least one gout- or osteoarthritis-related drug in addition to 2 diagnoses to minimize the exposure misclassification. For the main outcome, to maximize the specificity, we identified DM cases based on a diagnosis code and a dispensing for anti-diabetic drugs. Thus, patients with diet-controlled DM were not included. This may have resulted in an underestimation of the IR of DM in this study, as a prior study [10] that defined DM based on a fasting glucose level or use of anti-diabetic medication found higher IRs. Second, even though we adjusted for more than 40 variables in the multivariable Cox proportional hazards models, this study may be subject to residual confounding by family history of DM, physical activity, obesity, menopausal status among women, and diet. Third, the association in between gout and incident DM noted in this study may not be causal. However, our findings are consistent with a prior study [10] which reported a relative DM risk of 1.34 (95% CI 1.09–1.64) in men with gout after adjusting for age, body mass index, family history of DM, smoking, alcohol use, dietary factors, and presence of individual components of metabolic syndrome. Lastly, the mean followup period was less than 2 years. However, we expect similarly increased risks of DM associated with gout over time as the proportional hazards assumption was not violated in any of the models.
In conclusion, gout remained significantly associated with an increased risk of developing DM after adjusting for potential confounders, and the impact of gout on the risk of incident DM was higher among women than men. This study highlights the importance of patient education for lifestyle modification as well as careful monitoring of diabetic risk factors to reduce the risk of diabetes or metabolic syndrome among patients with gout, particularly women with gout. Because hyperuricemia and systemic inflammation may have a role in the causal pathway to type 2 DM, future research is needed to examine the role of aggressive gout management including urate-lowering therapy on the risk of DM.
Supplementary Material
Acknowledgments
Kim is supported by the NIH grant K23 AR059677. Kim received research support from Pfizer and tuition support for the Pharmacoepidemiology Program at the Harvard School of Public Health partially funded by the Pharmaceutical Research and Manufacturers of America (PhRMA) foundation.
Solomon is supported by the NIH grants K24 AR055989, P60 AR047782, and R01 AR056215. He serves in unpaid roles on studies sponsored by Pfizer, Novartis, Lilly, and Bristol Myers Squibb and receives royalties from UpToDate.com. Solomon received research support from Amgen, Pfizer and Lilly.
This study had no specific funding. Kim is supported by the NIH grant K23 AR059677. Solomon is supported by the NIH grants K24 AR055989, P60 AR047782, and R01 AR056215.
Footnotes
Disclosures:
Liu has nothing to disclose.
Competing interests
Kim receives research support from Pfizer and tuition support for the Pharmacoepidemiology Program at the Harvard School of Public Health partially funded by the Pharmaceutical Research and Manufacturers of America (PhRMA) foundation. Liu has nothing to disclose for financial support or conflict of interest. Solomon receives research support from Amgen, Lilly, Pfizer, and CORRONA, and serves in unpaid roles on studies sponsored by Pfizer, Novartis, Lilly, and Bristol Myers Squibb.
References
- 1.Choi HK, Mount DB, Reginato AM. Pathogenesis of gout. Annals of internal medicine. 2005 Oct 4;143:499–516. doi: 10.7326/0003-4819-143-7-200510040-00009. [DOI] [PubMed] [Google Scholar]
- 2.Zhu Y, Pandya B, Choi H. Prevalence of gout and hyperuricemia in the US general population: the National Health and Nutrition Examination Survey 2007–2008. Arthritis Rheum. 2011;63:3136–41. doi: 10.1002/art.30520. [DOI] [PubMed] [Google Scholar]
- 3.Choi HK, Ford ES. Prevalence of the metabolic syndrome in individuals with hyperuricemia. Am J Med. 2007 May;120:442–7. doi: 10.1016/j.amjmed.2006.06.040. [DOI] [PubMed] [Google Scholar]
- 4.Choi HK, Ford ES, Li C, et al. Prevalence of the metabolic syndrome in patients with gout: the Third National Health and Nutrition Examination Survey. Arthritis Rheum. 2007;57:109–15. doi: 10.1002/art.22466. [DOI] [PubMed] [Google Scholar]
- 5.Lai HM, Chen CJ, Su BY, et al. Gout and type 2 diabetes have a mutual inter-dependent effect on genetic risk factors and higher incidences. Rheumatology (Oxford) 2012 Apr;51:715–20. doi: 10.1093/rheumatology/ker373. [DOI] [PubMed] [Google Scholar]
- 6.Krishnan E, Akhras KS, Sharma H, et al. Relative and attributable diabetes risk associated with hyperuricemia in US veterans with gout. QJM: monthly journal of the Association of Physicians. 2013 Aug;106:721–9. doi: 10.1093/qjmed/hct093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Krishnan E, Pandya BJ, Chung L, et al. Hyperuricemia in young adults and risk of insulin resistance, prediabetes, and diabetes: a 15-year follow-up study. American journal of epidemiology. 2012 Jul 15;176:108–16. doi: 10.1093/aje/kws002. [DOI] [PubMed] [Google Scholar]
- 8.Sui X, Church TS, Meriwether RA, et al. Uric acid and the development of metabolic syndrome in women and men. Metabolism: clinical and experimental. 2008 Jun;57:845–52. doi: 10.1016/j.metabol.2008.01.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ryu S, Song J, Choi BY, et al. Incidence and risk factors for metabolic syndrome in Korean male workers, ages 30 to 39. Annals of epidemiology. 2007 Apr;17:245–52. doi: 10.1016/j.annepidem.2006.10.001. [DOI] [PubMed] [Google Scholar]
- 10.Choi HK, De Vera MA, Krishnan E. Gout and the risk of type 2 diabetes among men with a high cardiovascular risk profile. Rheumatology (Oxford) 2008 Oct;47:1567–70. doi: 10.1093/rheumatology/ken305. [DOI] [PubMed] [Google Scholar]
- 11.De Vera MA, Rahman MM, Bhole V, et al. Independent impact of gout on the risk of acute myocardial infarction among elderly women: a population-based study. Ann Rheum Dis. 2010 Jun;69:1162–4. doi: 10.1136/ard.2009.122770. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Grayson PC, Kim SY, LaValley M, et al. Hyperuricemia and incident hypertension: a systematic review and meta-analysis. Arthritis Care Res (Hoboken) 2011 Jan;63:102–10. doi: 10.1002/acr.20344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kim SY, Guevara JP, Kim KM, et al. Hyperuricemia and coronary heart disease: a systematic review and meta-analysis. Arthritis Care Res (Hoboken) 2010 Feb;62:170–80. doi: 10.1002/acr.20065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Puig JG, Michan AD, Jimenez ML, et al. Female gout. Clinical spectrum and uric acid metabolism. Arch Intern Med. 1991 Apr;151:726–32. doi: 10.1001/archinte.151.4.726. [DOI] [PubMed] [Google Scholar]
- 15.Lally EV, Ho G, Jr, Kaplan SR. The clinical spectrum of gouty arthritis in women. Arch Intern Med. 1986 Nov;146:2221–5. [PubMed] [Google Scholar]
- 16.Meisinger C, Thorand B, Schneider A, et al. Sex differences in risk factors for incident type 2 diabetes mellitus: the MONICA Augsburg cohort study. Arch Intern Med. 2002 Jan 14;162:82–9. doi: 10.1001/archinte.162.1.82. [DOI] [PubMed] [Google Scholar]
- 17.Gagne JJ, Glynn RJ, Avorn J, et al. A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64:749–59. doi: 10.1016/j.jclinepi.2010.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rothman K, Greenland S, Lash T. Modern Epidemiology. 3. Philadelphia: Lippincott Williams & Wilkins; 2008. [Google Scholar]
- 19.Cox D. Regression models and life-tables. J R Stat Soc. 1972;34:187–220. [Google Scholar]
- 20.Kleinbaum D, Klein M. Evaluating the Proportional Hazards Assumption. In: Gail M, Krickberg K, Samet J, et al., editors. Survival Analysis: A Self-Learning Text. 3. Springer; 2012. [Google Scholar]
- 21.Carnethon MR, Fortmann SP, Palaniappan L, et al. Risk factors for progression to incident hyperinsulinemia: the Atherosclerosis Risk in Communities Study, 1987–1998. American journal of epidemiology. 2003 Dec 1;158:1058–67. doi: 10.1093/aje/kwg260. [DOI] [PubMed] [Google Scholar]
- 22.Niskanen L, Laaksonen DE, Lindstrom J, et al. Serum uric acid as a harbinger of metabolic outcome in subjects with impaired glucose tolerance: the Finnish Diabetes Prevention Study. Diabetes care. 2006 Mar;29:709–11. doi: 10.2337/diacare.29.03.06.dc05-1465. [DOI] [PubMed] [Google Scholar]
- 23.Festa A, D’Agostino R, Jr, Tracy RP, et al. Elevated levels of acute-phase proteins and plasminogen activator inhibitor-1 predict the development of type 2 diabetes: the insulin resistance atherosclerosis study. Diabetes. 2002 Apr;51:1131–7. doi: 10.2337/diabetes.51.4.1131. [DOI] [PubMed] [Google Scholar]
- 24.Pradhan AD, Manson JE, Rifai N, et al. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA: the journal of the American Medical Association. 2001 Jul 18;286:327–34. doi: 10.1001/jama.286.3.327. [DOI] [PubMed] [Google Scholar]
- 25.Herder C, Peltonen M, Koenig W, et al. Systemic immune mediators and lifestyle changes in the prevention of type 2 diabetes: results from the Finnish Diabetes Prevention Study. Diabetes. 2006 Aug;55:2340–6. doi: 10.2337/db05-1320. [DOI] [PubMed] [Google Scholar]
- 26.Schmidt MI, Duncan BB, Sharrett AR, et al. Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities study): a cohort study. Lancet. 1999 May 15;353:1649–52. doi: 10.1016/s0140-6736(99)01046-6. [DOI] [PubMed] [Google Scholar]
- 27.Shelmadine B, Bowden RG, Wilson RL, et al. The effects of lowering uric acid levels using allopurinol on markers of metabolic syndrome in end-stage renal disease patients: a pilot study. Anadolu kardiyoloji dergisi: AKD = the Anatolian journal of cardiology. 2009 Oct;9:385–9. [PubMed] [Google Scholar]
- 28.Feig DI, Soletsky B, Johnson RJ. Effect of allopurinol on blood pressure of adolescents with newly diagnosed essential hypertension: a randomized trial. JAMA: the journal of the American Medical Association. 2008 Aug 27;300:924–32. doi: 10.1001/jama.300.8.924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Terawaki H, Nakayama M, Miyazawa E, et al. Effect of allopurinol on cardiovascular incidence among hypertensive nephropathy patients: the Gonryo study. Clinical and experimental nephrology. 2012 Nov 29; doi: 10.1007/s10157-012-0742-z. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.