Abstract
Objective. To examine the incidence of gout over the last 20 years and to evaluate possible changes in associated comorbid conditions.
Methods. The medical records were reviewed of all adults with a diagnosis of incident gout in Olmsted County, Minnesota, USA, during 2 time periods (January 1, 1989–December 31, 1992, and January 1, 2009–December 31, 2010). Incident cases had to fulfill at least 1 of 3 criteria: the American Rheumatism Association 1977 preliminary criteria for gout, the Rome criteria, or the New York criteria.
Results. A total of 158 patients with new-onset gout were identified during 1989–1992 and 271 patients during 2009–2010, yielding age- and sex-adjusted incidence rates of 66.6/100,000 (95% CI 55.9–77.4) in 1989–1992 and 136.7/100,000 (95% CI 120.4–153.1) in 2009–2010. The incidence rate ratio was 2.62 (95% CI 1.80–3.83). At the time of their first gout flare, patients diagnosed with gout in 2009–2010 had higher prevalence of comorbid conditions compared with 1989–1992, including hypertension (69% vs 54%), diabetes mellitus (25% vs 6%), renal disease (28% vs 11%), hyperlipidemia (61% vs 21%), and morbid obesity (body mass index ≥ 35 kg/m2; 29% vs 10%).
Conclusion. The incidence of gout has more than doubled over the recent 20 years. This increase together with the more frequent occurrence of comorbid conditions and cardiovascular risk factors represents a significant public health challenge.
Gout is the most common inflammatory arthritis in the United States1. Several studies have demonstrated an increasing prevalence over recent decades2,3,4,5. The number of people living with disability due to gout has increased dramatically over the last 20 years6.
Whether there has been an actual increase in the incidence of gout in recent years is unclear. Previous population-based studies using the Rochester Epidemiology Project (REP) have suggested that the incidence of gout nearly doubled between 1977–1978 and 1995–19964, but population-based data about gout incidence in the United States over the last 20 years are limited7,8.
While the morbidity associated with gout itself is worrisome, patients with gout also have a higher likelihood of various comorbid conditions such as obesity, hyperlipidemia, renal disease, and hypertension (HTN)9. Possible changes in the prevalence of these comorbidities among individuals with gout would have a tremendous effect on the overall morbidity of patients with gouty arthritis.
The purpose of our present study was to examine trends in the incidence of gout in a US-based population, and to determine whether the prevalence of some comorbid conditions having major effects on patient well-being and survivorship among affected patients has changed.
MATERIALS AND METHODS
Olmsted County, Minnesota, USA, is well suited for population-based studies through the resources of the REP10. The REP is a medical record linkage system that enumerates the population and provides ready access to the medical records from all healthcare providers in Olmsted County, Minnesota, USA, including the Mayo Clinic, the Olmsted Medical Center and its affiliated hospitals, local nursing homes, and the few private practitioners. The REP ensures the full identification of all the medical records, both inpatient and outpatient, for the population. The REP ensures virtually complete ascertainment of all clinically recognized cases of gout among the residents of Olmsted County, Minnesota. This study was approved by the institutional review boards of the Mayo Clinic (12-007239) and the Olmsted Medical Center (018-OMC-15).
Using the REP, adult (age ≥ 18 yrs) Olmsted County residents with the potential diagnosis of gout were retrieved using diagnostic codes (International Classification of Diseases, 9th ed: code 274.x) during 2 different time periods (January 1, 1989–December 31, 1992, and January 1, 2009–December 31, 2010). The medical records (inpatient and outpatient) of each potential case were then reviewed for ascertainment of gout diagnosis. Cases were identified as incident at the earliest date the case fulfilled at least 1 of 3 proposed gout criteria: the American Rheumatism Association (ARA) 1977 preliminary criteria for gout11, the Rome criteria12, or the New York criteria13 (Table 1).
Criteria used for gout diagnosis: 1977 ARA preliminary criteria, Rome criteria, and New York criteria.
Demographic and clinical data were abstracted through medical record review. Smoking status (never, current, former) at the time of the incident gout diagnosis was collected. Body mass index (BMI) was calculated using the standing height and weight recorded in the medical record at the time of gout diagnosis. BMI was categorized based on the World Health Organization classification: normal (18.5–24.9 kg/m2), overweight (pre-obese; 25–29.9 kg/m2), class I obesity (30–34.9 kg/m2), class II obesity (35–39.9 kg/m2), class III obesity (≥ 40 kg/m2)14. Cardiovascular (CV) and metabolic risk factors and comorbidities were abstracted from the records based on physician diagnoses and laboratory values (i.e., lipid profile) prior to or at the diagnosis of gout. Concomitant medication use for CV risk factors and comorbidities were recorded.
Statistical analysis
Descriptive statistics (means, percentages, etc.) were used to summarize data. Clinical characteristics were compared between the time periods using Wilcoxon rank sum tests and chi-square tests. P values < 0.05 were considered statistically significant. Age- and sex-specific incidence rates were calculated using the number of incident cases as the numerator and population estimates for adults (age ≥ 18 yrs) based on decennial census counts as the denominator, with linear interpolation used to estimate population size for intercensal years. Incidence rates were age- and sex-adjusted for the US 2010 white population. To compute 95% CI for incidence rates, it was assumed that the number of incident cases followed a Poisson distribution. Incidence rates were compared between the 2 time periods using Poisson regression methods. Analyses were performed using SAS version 9.4 (SAS Institute) and R 3.1.1 (R Foundation for Statistical Computing).
RESULTS
A total of 429 patients with incident gout were identified: 158 in 1989–1992 and 271 in the 2009–2010 time period. While the age and sex distribution of individuals diagnosed with gout remained similar when comparing both time periods (Table 2), the proportion of nonwhites increased from 6% to 10%. BMI in the 2009–2010 cohort was significantly higher than that of the 1989–1992 cohort (median 30.9 vs 28.3, respectively; p < 0.001). This increase in BMI was observed in both men and women. Moreover, the number of individuals with class II and III obesity tripled between the earlier and the later cohort (29% vs 10% with BMI ≥ 35 kg/m2 in 2009–2010 vs 1989–1992, respectively; p < 0.001).
Patient characteristics, major associated comorbidities, and medication use among Olmsted County, Minnesota, USA residents with incident gout in 1989–1992 compared with 2009–2010. Values are n (%) unless otherwise specified.
In addition, there were notably higher rates of CV risk factors and comorbidities in the 2009–2010 cohort, including hyperlipidemia, diabetes mellitus (DM), HTN, and renal disease (Table 2). The use of aspirin increased from 23% to 39% between the 1989–1992 and 2009–2010 cohorts (p < 0.01), and the use of statins increased from 4% to 41% between these 2 periods. Of note, the serum levels of total and low-density lipoprotein cholesterol were significantly lower in the 2009–2010 cohort than the 1989–1992 cohort, while high-density lipoprotein levels were higher in the more recent cohort.
Using the combined criteria set, the age- and sex-adjusted incidence rate of gout among adults (age ≥ 18 yrs) in the 1989–1992 time period was 66.6/100,000 (95% CI 55.9–77.4) compared with 136.7/100,000 (95% CI 120.4–153.1) in the 2009–2010 time period. This constitutes a more than 2-fold increase of incident gout within 20 years (rate ratio 2.62, 95% CI 1.80–3.83; p < 0.001). This increase affected both sexes to a similar extent (Figure 1) and involved all age groups (Table 3).
Incidence of gout among adult (age ≥ 18 yrs) Olmsted County, Minnesota, USA residents in 1989–1992 and 2009–2010, based on the earliest date of fulfillment of the 1977 American Rheumatism Association, Rome, or New York criteria, according to age and sex.
Incidence of gout among adult (age ≥ 18 years) Olmsted County, Minnesota residents in 1989–1992 and 2009–2010 based on the earliest date of 1977 American Rheumatism Association, Rome, or New York criteria, according to age and sex.
The significant increase in the incidence of gout was also apparent when applying each of the criteria separately. Using the 1977 ARA preliminary criteria identified 118 incident cases in 1989–1992 and 183 incident cases in 2009–2010; the age- and sex-adjusted incidence rate rose from 49.5 (95% CI 40.2–58.7) in the 1989–1992 cohort to 92.7 (95% CI 79.2–106.2) in the 2009–2010 cohort (rate ratio 2.64, 95% CI 1.69–4.13). When the New York criteria were used, 91 incident cases in 1989–1992 and 217 incident cases in 2009–2010 were identified and the age- and sex-adjusted incidence rate increased from 39.0 (95% CI 30.7–47.2) in the 1989–1992 cohort to 110.0 (95% CI 95.2–124.7) in the 2009–2010 cohort (rate ratio 3.41, 95% CI 2.13–5.44). Finally, applying the Rome criteria (119 incident cases in 1989–1992 and 240 incident cases in 2009–2010) resulted in an age- and sex-adjusted incidence rate of 50.4 (95% CI 41.0–59.7) in the 1989–1992 cohort and 121.3 (95% CI 105.9–136.8) in the 2009–2010 cohort (rate ratio 2.94; 95% CI 1.95–4.43).
The serum uric acid level at the time of the first attack was not significantly different between the cohorts (median 8.0 in the 1989–1992 vs 8.2 in the 2009–2010 cohort; p = 0.65; Table 2). Similarly, the use of diuretics at the time of incident gout did not differ significantly (43% in the 1989–1992 vs 48% in the 2009–2010 cohort; p = 0.32).
DISCUSSION
Our population-based study demonstrates a marked increase in the incidence of gout among residents of Olmsted County, Minnesota, over the last 20 years. The rate of newly diagnosed cases of gout more than doubled within the observed time period for both men and women and across all age groups. Importantly, these results supplement numerous studies that have demonstrated an increase in the prevalence of gout, not only in the United States but in various geographic settings1–5,15–20.
Why are the incidence and prevalence rates of gout increasing at such a remarkable rate? Our study was not designed to investigate the underlying cause for this increase, but other authors have suggested changes in risk factors of gout as the main contributing factor9,16. Risk factors of gout that have been increasing in the US population over recent years include obesity, age, DM, chronic renal disease, and aspirin use21,22,23,24,25,26,27,28. Moreover, as the population advances in age, serum uric acid level was found to increase as well29. Interestingly, despite the significant increases in diagnosis of HTN, there were no differences in the use of thiazide diuretics noted between time periods. Thus, increased thiazide diuretic use is unlikely to be responsible for the increasing incidence of gout in this population, despite previous reports linking them to gout occurrence30.
There has been a marked increase in the prevalence of CV comorbidities among patients with gout. In our study, diseases including obesity, HTN, DM, chronic renal disease, and hyperlipidemia have become significantly more common in patients diagnosed with gout in the more recent years of 2009–2010 compared with those diagnosed with gout 20 years ago. The change in the prevalence of obesity in patients with gout is particularly noteworthy. Severe, class II, and class III obesity have tripled in this cohort over the past 20 years.
The high prevalence of comorbidities among patients with gout has been demonstrated in several previously published cohorts21,22,28,31. For example, in the 2007–2008 US National Health and Nutrition Examination Survey cohort, Zhu, et al found a prevalence of 74% HTN, 71% chronic renal disease, 53% obesity, 26% DM, and 11% heart failure among patients with gout22. Except for chronic renal disease, which was less common in our study, these numbers are comparable to the findings in the 2009–2010 cohort reported in our current study.
Our study has a number of strengths. The population-based approach and availability of the complete medical records for all patients offer the advantage of a comprehensive assessment of all patients diagnosed with gout in a defined geographic area, thereby minimizing the risk of referral bias. The use of a systematic approach to case identification using multiple criteria sets in both time periods is another strength of the study.
Inherent limitations of the study relate to its retrospective design. The assessment and documentation of criteria used to diagnose gout may have changed over time. For example, arthrocentesis with crystal analysis was more commonly performed in more recent years. This may reflect a change in practice patterns. Use of all 3 proposed criteria sets for gout mitigates against imprecision of case ascertainment11,12,13, lowering the dependence on single criteria that may have been assessed or documented differently over time. In addition, a subanalysis was performed using each criteria set independently, with each showing a significant increase in the incidence of gout over time. Since the completion of our study, new gout classification criteria have been published32. These criteria incorporate new imaging modalities including musculoskeletal ultrasound and dual-beam computed tomography. Because these imaging methods were not used during the 1989–1992 time period, but were widely available during the 2009–2010 period, the 2015 American College of Rheumatology/European League Against Rheumatism classification criteria were not applied to our current study, to reduce imbalance in diagnostic assessment resulting in ascertainment bias.
The classification of comorbidities was based on physician diagnosis as reflected in the medical record, therefore the possibility of changing documentation patterns over time could affect ascertainment. However, this would not apply to diagnoses of obesity, because classification was based on BMI, and weight and height have been documented for every patient in both cohorts. Finally, the population of Olmsted County is mainly white and the socioeconomic characteristics largely resemble those of the US white population in general33. Caution has to be exercised when generalizing these findings to populations with a different demographic composition, especially in other populations with higher prevalence of African Americans; higher incidence rates have been previously reported in the latter population compared to white populations34,35,36.
Our data demonstrate a significant increase in the incidence of gout over the last 20 years. Gout is the most common inflammatory arthritis in the United States and affects more than 3% of the population1. This rise in incidence is combined with an alarming increase in comorbidities such as HTN, hyperlipidemia, DM, renal disease, and a tripling of grades II and III obesity among patients with gout. The importance of these comorbidities cannot be overemphasized; previous studies have demonstrated their significant contribution to the excess mortality and morbidity in patients with gout37.
These findings emphasize the importance of acknowledging gout as a “sentinel” disease that rarely occurs in isolation, but points to a likely aggregation of various CV risk factors and comorbidities. Thus, in most patients, management of the initial gout flare will represent only a minor component of treatment. A comprehensive, multispecialty approach is needed to reduce the morbidity and mortality of gout and its associated health conditions in these patients.
Footnotes
This work was made possible using the resources of the Rochester Epidemiology Project, which is supported by the US National Institute on Aging of the National Institutes of Health (NIH) under award number R01AG034676 and Clinical and Translational Science Awards grant number UL1 TR000135 from the National Center for Advancing Translational Sciences, a component of the NIH.
- Accepted for publication October 18, 2017.
