Wahedduddin, Salman MBBS*; Singh, Jasvinder A. MBBS, MPH*†; Culhane-Pera, Kathleen A. MD, MA‡; Gertner, Elie MD*§
Gout has been reported to be more prevalent in certain ethnicities, including the Maori in New Zealand1 and aboriginal Chinese in Taiwan.2 These groups have higher serum uric acid (SUA) levels as compared with controls and gouty arthritis is more prevalent in these ethnic groups compared with other populations.2,3 The prevalence of other risk factors does not seem to differ much from controls. Because hyperuricemia has been shown to be the most important risk factor leading to gout in these groups, there needs to be more awareness of the importance of treating gout in these populations.3,4
Hmong are an ethnic group originally from southern China that later migrated to the northern regions of Laos, Vietnam, and Thailand. Since the 1970s, large numbers of Hmong have come to the United States as refugees.5 It is estimated that there are 190,000 Hmong in the United States, with Minnesota having the second largest Hmong population after California.5 The estimates of Hmong in Minnesota vary from 42,000 to 65,000 with the Twin Cities area having the largest metropolitan concentration of Hmong in the country.5,6 Many physicians suspect that Hmong have more tophaceous gout as compared with Whites.7 Hmong have increased risk for uric acid stones compared with controls.8 The Hmong in the United States have an increased risk and earlier development of obesity, hypertension, and diabetes compared with Whites, factors that increase the risk of hyperuricemia and possibly gout.9,10
There are no previous large studies of gout in the Hmong, or whether the risk factors for gout in Hmong are similar to those reported earlier in whites.7
Our aim in this study was to explore the disease characteristics and risk factors for gout in the Hmong patients compared with Whites. We hypothesized that in our sample: (a) Hmong have more severe gout (more tophi, more flares, higher SUA) than Whites; (b) this increased risk of tophaceous gout persists even after adjusting for known risk factors for gout including age, sex, kidney disease, hypertension, and diuretic use; and (c) that Hmong get less aggressive treatment despite earlier and more severe presentation, compared with their white counterparts.
MATERIALS AND METHODS
Using the HealthPartners clinical databases in St. Paul, Minnesota, we performed a retrospective chart review of Hmong and Whites with at least 1 International Classification of Diseases, ninth revision code for gout, during the period January 1, 2001 to March 31, 2008. Electronic records were available from 2001 for patients in the health system's 36 clinics. Since the primary aim of this study was descriptive, we searched from January 1, 2001 to March 31, 2008 to ensure a maximum number of patients studied.
Since documented codes can be inaccurate, we used the presence of physician-documented gout diagnoses on 2 or more occasions at least 4-weeks apart in the chart to confirm the diagnosis of gout. These patients constituted the analytic dataset. Earlier studies have used this criterion for other rheumatologic diseases with good accuracy.11,12 We also examined which patients in the analytic set had the preliminary American College of Rheumatology (ACR) classification criteria for gout.13
Individuals of Hmong ethnicity have 1 of 18 unique clan names. We searched for Hmong patients using these clan names and for documentation of the Hmong language being the main language to identify all Hmong patients from the database. This approach has been validated for descriptive studies of Hmong populations.8,14 We selected the White controls by using the most common White last names in Minnesota of “Anderson” and “Johnson.” White ethnicity was confirmed by documentation in the electronic medical records. Patient-reported ethnicity was used to define the White group.
We used a standardized data extraction form to obtain the following data recorded in provider notes, pharmacy prescriptions, laboratory databases, and administrative/billing databases. Data were extracted by a trained rheumatology fellow (S.W.) for the entire duration of follow-up at HealthPartners, starting from the initial visit for gout until the last note or March 31, 2008, whichever occurred later.
The following data were extracted: age at onset of symptoms, number and location of joints and tophi, presence of >3 attacks/yr, all SUA levels, glomerular filtration rates, allopurinol dose and duration, and use of other drugs to treat gout including colchicine, uricosuric drugs, non-steroidal anti-inflammatory drugs (NSAIDs), and corticosteroids. We also recorded age, height, and weight to calculate body mass index (BMI), family history of gout, and alcohol use (heavy alcohol usage was defined as greater than 7 drinks per week). When baseline alcohol use data were missing, we used the data from the earliest point of documentation as a surrogate for baseline. For missing height data on patients with weight documentation, mean heights for each ethnicity were used to calculate BMI (8 Hmong and 2 Whites). Comorbidities such as hypertension, renal stones, diabetes, hyperlipidemia, and coronary artery disease were recorded. We recorded whether these were present at the time of onset of gout symptoms, at the time of physician diagnosis, and anytime during the follow-up. We recorded number and name of joints involved as right, left, or bilateral involvement at presentation and cumulative number of joints involved at initial attack and for the duration of follow-up.
A patient was defined as having tophaceous gout if the examining physician documented the presence of tophi. Gouty attack was defined as documentation of acute onset of joint pain with some clinical features of inflammation, or treatment in keeping with that used for an acute flare, including NSAIDs, colchicine, or prednisone.
Clinical characteristics were compared between the Hmong and Whites using χ2 and t tests, as appropriate. Univariate and multivariable-adjusted logistic regressions compared the association of Hmong ethnicity as the main predictor of interest and presence of tophaceous gout as the outcome of interest. Hierarchical regression models controlled for the following variables: (1) Model 1 included age, sex, and ethnicity; (2) Model 2 included age, sex, ethnicity, hypertension, and diuretic use (Model 1 + hypertension + diuretic use). Sensitivity analyses were conducted by restricting these analyses to the subset that met the ACR criteria (35%, 18%, 40%, and 39%, respectively), which would likely lead to further selection bias and potential for reporting/recall bias for symptom duration and alcohol use.
We identified 136 Hmong and 146 Whites with an International Classification of Diseases, ninth revision code for gout from the database. About 25 Hmong patients and 32 Whites were excluded because of incorrect or missing ethnicity in the chart. This provided us with 111 Hmong and 114 whites with confirmed ethnicity. About 89 Hmong (80%) and 84 Whites (74%) met the 2 physician diagnoses criteria for gout and constituted the analytic cohort for most analyses. Of these, 61 Hmong and 47 Whites met the preliminary ACR criteria for gout used for sensitivity analyses (Fig. 1).
The Hmong group was younger (58.3 vs. 66.3 years, P = 0.04), and consisted of more males (84.6% vs. 77.4%, P = 0.25). Hmong had lower prevalence of diabetes mellitus, coronary vascular disease, and hyperlipidemia than Whites at initial presentation (Table 1). The Hmong were leaner (BMI, 28.3 vs. 32.9 P = < 0.001), although the BMI data were only available for 53 Hmong and 51 White. About 20% Whites and 18% Hmong had monosodium urate crystal identification. Only 18% of Whites had seen a rheumatologist as opposed to 30% Hmong (P = 0.06).
Gout Severity in Hmong Versus Whites
Hmong had a much earlier onset of symptoms (37.4 vs. 55 years, P < 0.001) and higher mean SUA levels during follow-up compared with Whites (9.1 vs. 7.6 mg/dL, P ≤ 0.001) (Table 2). These results were unchanged when analyses were restricted to the ACR criteria subset. Surprisingly, 38.5% of Hmong presented with gout symptoms before the age of 30 as opposed to only 8.2% of Whites. Hmong had more joint involvement (3 vs. 2.5, P = 0.014), renal calculi (11.2% vs. 2.4% P = 0.023), and increased frequency of acute attacks per year (45.9% vs. 28.9%, P = 0.027). There was no significant difference in the duration of gout symptoms between the Hmong and Whites (10.5 vs. 11.4 years, P = 0.6). This result did not change when restricting the analysis to the ACR criteria subset (Table 2).
Treatment of Gout in the Groups
Allopurinol was prescribed more often in the Hmong (74.2% vs. 59.5%, P = 0.04), however, the Hmong used allopurinol for fewer months than the Whites (30 vs. 48, P = 0.05). These differences were not detectable in the ACR subgroup (Table 3). There was a difference in allopurinol mean dose with the Hmong using 241 mg/d and the White group using 341 mg/d (P = 0.1). The Hmong had fewer gout-related outpatient visits with a diagnostic code of gout as compared with the Whites (7.1 vs. 5.8, P = 0.012).
Univariate and Multivariable Analyses of Ethnicity as Risk Factor for Tophaceous Gout
Hmong were more likely to have tophaceous gout than Whites (38.9% vs. 15.1%, P = 0.004) (Table 2). The mean number of tophi was 1.0 in Hmong, as opposed to only 0.25 in Whites (P < 0.001). In the ACR subgroup, Hmong also had more tophi (1.21 vs. 0.32, P = 0.001). In addition, Hmong had more hand tophi (21.3% vs. 3.6%, P < 0.001). This difference was also significant in the ACR subgroup (27.9% vs. 6.4%, P = 0.004). Multivariable analyses adjusted for age and gender found a significant association of Hmong ethnicity with risk of tophaceous gout (Table 4). The association was significant after adjusting additionally for hypertension and diuretic use.
Hmong are thought to originate from Southern Chinese aboriginal population.15,16 In this study, we examined if Hmong have more severe gout compared with Whites, using retrospective data. Our study made several interesting observations, which deserve further discussion.
The most striking observation was that the Hmong group had an earlier onset of gout symptoms. More than one-third of Hmong patients were less than 30 years old at presentation as compared with only 8% of Whites. This earlier disease onset in Hmong is less likely to be due to detection bias due to physician visits for other conditions, as these patients had lower comorbidity load compared with Whites. Hmong also had increased number of tophi especially hand tophi and higher mean SUA at disease onset and during follow-up. Increased tophi, renal calculi, higher number of acute gout attacks, and more joints involved are all manifestations of related to prolonged duration of elevated SUA and/or lower rates of allopurinol use in the Hmong. Some well-known risk factors for gout were far less prevalent in Hmong compared with our White group including hypertension, chronic kidney disease, diuretic use, obesity, and heavy alcohol use. Thus, at least the traditional risk factors do not explain the risk of earlier onset or more severe gout in the Hmong. Generally, the comorbidity load is expected to less in a younger group and this could explain the lack of traditional risk factors.
Hmong patients used allopurinol for fewer months, even though they were prescribed allopurinol more often and had fewer visits for gout, despite having more frequent attacks and more tophi. The cumulative allopurinol dose was also lower than Whites. Thus, medication compliance and follow-up in Hmong population are challenges to successful gout treatment and management. Many factors such as barriers of language, access, documentation, and socioeconomic factors are likely to impact access and awareness of risks/benefits of treatment.3 Cultural disease models and culture-based alternative treatment can also impact effectiveness of care of chronic disease.3 Treatment in the White group was also inadequate as their SUA was greater then the goal of 6 mg/dL. This is in accordance with the inadequate treatment of gout in general in the United States.17
Hmong ethnicity was associated with tophaceous gout after controlling for age, and gender, hypertension, and diuretic use. We could not control for duration of gout, BMI, glomerular filtration rate, and alcohol use because of missing data in these variable to control for further bias. Another reason for not controlling for gout symptom duration is that this is on the pathway to tophaceous gout and adjusting for it would lead to over-adjustment in models. Hmong may have genetic predisposition to hyperuricemia similar to Polynesian or South Chinese groups, a hypothesis that needs further study.3 It is possible that decreased urinary urate excretion as noted in the Polynesian population may also be present in the Hmong population. GLUT9 polymorphism affecting urate under-excretion was recently identified in New Zealand's Maori population.18 Genetic susceptibility loci,19,20 including polymorphisms in Hypoxanthine guanine phosphoribosyltransferase,21 have also been reported in other studies of Taiwanese aboriginal families. Since all Hmong have a limited set of clan names they are most likely related closely, and thus there may be a genetic risk factor Dietary differences between US Hmong and Whites may also be responsible for the higher prevalence of tophaceous gout noted in the Hmong.10,22
Our study has several limitations. Retrospective study design makes our findings susceptible to residual confounding bias. Documentation of BMI and alcohol use was missing in almost half of the patients. Duration of gout symptoms was not available in 40% of the Hmong. Quantitative alcohol consumption was rarely documented unless it was excessive, limiting the ability to see the effect of this important risk factor.
This study has several strengths. We had almost 70 to 80 patients in each group to perform most comparisons. The main results of our study were robust, since they were confirmed when we restricted the analyses to patients meeting the ACR classification criteria. Although it is challenging to find documentation of ACR criteria in database studies,23 this approach is considered valid by most researchers. Crystal identification, which is considered the gold standard, is not frequently available and clinical criteria have reasonable predictive value.24 Previous studies also found that physician documentation of rheumatic diseases (rheumatoid arthritis and spondyloarthropathies) in medical records on 2 or more occasions agreed well with ACR criteria, another definition we used in our study.11,12
In summary, Hmong need aggressive treatment of their gout because of their tendency to have earlier presentation, tophaceous gout with more frequent acute attacks, and hand tophi which can be particularly disabling and predict poor functional capability.25 Increased duration of untreated or undertreated gout will lead to more severe clinical gout. Control of gout with close follow-up can prevent the chronic manifestations.26 Further population-based studies may be necessary to determine if Hmong have similar prevalence and genetic risk factors for hyperuricemia as compared with other Aboriginal Chinese and Polynesian populations.
The authors thank Teresa Defor for extracting data from the HealthPartners database, Steven Asche for valuable help with IRB submission, and The HealthPartners Research Foundation, Institute of Medical Education for their support and grant.
1. Gibson T, Waterworth R, Hatfield P, et al. Hyperuricaemia, gout and kidney function in New Zealand Maori men. Br J Rheumatol
2. Chou CT, Lai JS. The epidemiology of hyperuricaemia and gout in Taiwan aborigines. Br J Rheumatol
3. Padang C, Muirden KD, Schumacher HR, et al. Characteristics of chronic gout in Northern Sulawesi, Indonesia. J Rheumatol
4. Winnard D, Kake T, Gow P, et al. Debunking the myths to provide 21st century management of gout. N Z Med J
5. Pfeifer ME, Lee S. Hmong population, demographic, socioeconomic and educational trends in the 2000 census. In: Hmong 2000 Census Publication: Data and Analysis
.Saint Paul, MN: Hmong National Development, Inc. & Hmong Cultural and Resource Center; 2004.
6. Ronningen BJ. Estimates of Selected Immigrant Populations in Minnesota
. Minneapolis, MN: Minnesota State Demographic Center; 2004. Publication No. 05D-04–112.3.
8. Portis AJ, Hermans K, Culhane-Pera KA, et al. Stone disease in the Hmong of Minnesota: initial description of a high-risk population. J Endourol
9. Her C, Mundt M. Risk prevalence for type 2 diabetes mellitus in adult Hmong in Wisconsin: a pilot study. Wis Med J
10. Stang J, Kong A, Story M, et al. Food and weight-related patterns and behaviors of Hmong adolescents. J Am Diet Assoc
11. Singh JA, Holmgren AR, Krug H, et al. Accuracy of the diagnoses of spondylarthritides in veterans affairs medical center databases. Arthritis Rheum
12. Singh JA, Holmgren AR, Noorbaloochi S. Accuracy of Veterans Administration databases for a diagnosis of rheumatoid arthritis. Arthritis Rheum
13. Wallace SL, Robinson H, Masi AT, et al. Preliminary criteria for the classification of the acute arthritis of primary gout. Arthritis Rheum
14. Mills PK, Yang R. Cancer incidence in the Hmong of Central California, United States, 1987–94. Cancer Causes Control
15. Wen B, Li H, Gao S, et al. Genetic structure of Hmong-mien speaking populations in East Asia as revealed by mtDNA lineages. Mol Biol Evol
16. Li H, Cai X, Winograd-Cort ER, et al. Mitochondrial DNA diversity and population differentiation in southern East Asia. Am J Phys Anthropol
17. Singh JA, Hodges JS, Toscano JP, et al. Quality of care for gout in the US needs improvement. Arthritis Rheum
18. Dalbeth N, Hollis-Moffat JE, Xu X, et al. The GLUT9
gene and susceptibility to gout in the New Zealand population. Arthritis Rheum
. In press.
19. Cheng LS, Chiang SL, Tu HP, et al. Genomewide scan for gout in Taiwanese aborigines reveals linkage to chromosome 4q25. Am J Hum Genet
20. Wang WH, Chang SJ, Wang TN, et al. Complex segregation and linkage analysis of familial gout in Taiwanese aborigines. Arthritis Rheum
21. Chang SJ, Chang JG, Chen CJ, et al. Identification of a new single nucleotide substitution on the hypoxanthine-guanine phosphoribosyltransferase gene (HPRTTsou) from a Taiwanese aboriginal family with severe gout. J Rheumatol
22. Choi HK, Atkinson K, Karlson EW, et al. Purine-rich foods, dairy and protein intake, and the risk of gout in men. N Engl J Med
23. Harrold LR, Saag KG, Yood RA, et al. Validity of gout diagnoses in administrative data. Arthritis Rheum
24. Malik A, Schumacher HR, Dinnella JE, et al. Clinical diagnositic criteria for gout: comparison with the gold standard of synovial fluid crystal analysis. J Clin Rheumatol
25. Dalbeth N, Collis J, Gregory K, et al. Tophaceous joint disease strongly predicts hand function in patients with gout. Rheumatology
26. Darmawan J, Rasker JJ, Hendri N. The effect of control and self-medication on chronic gout in a developing country: outcome after 10 years. J Rheumatol
© 2010 Lippincott Williams & Wilkins, Inc.