The epidemiology of hyperuricaemia and gout in Taiwan aborigines.
BRITISH JOURNAL OF RHEUMATOLOGY
THE EPIDEMIOLOGY OF HYPERURICAEMIA AND GOUT IN TAIWAN ABORIGINES
C. T. CHOU 0 1
J. S. LAI 0 1 2
0 Submitted 24 February 1997; revised version accepted 26 June 1997. Immunology, China Medical College Hospital , No. 75, Yuh-Der Road, Taichung , Taiwan
1 Division of Rheumatology-Immunology, Department of Medicine, China Medical College Hospital
2 Department of Public Health, China Medical College , Taichung , Taiwan
To determine the prevalence of hyperuricaemia, gout and gout-related factors in Central Taiwan Atayal aborigines, 342 subjects over 18 yr old were interviewed and examined. A questionnaire was designed to screen for signs and symptoms of gout and gout-related risk factors. Serum uric acid, triglyceride and creatinine were measured in all subjects. The prevalence of hyperuricaemia was 41.4% and that of gout 11.7% in aborigines. The uric acid level was 7.9 ± 1.7 mg/dl in males and 5.7 ± 1.5 in females, and diVered significantly under age 70 yr (P < 0.001). Significantly increased triglyceride, creatinine and alcoholism was found in gouty patients compared with non-gouty patients. In 40 cases with gout, 54% had tophi and 35% of their firstdegree relatives had gout. The high prevalence of hyperuricaemia and gout in Taiwan Atayal aborigines, a significant family predisposition, increased creatinine level and alcoholism suggest multiple factors aVecting the hyperuricaemia.
Epidemiology; Atayal aborigine; Risk factors; Hyperuricaemia; Gout; Tophi
Depending on its case definition, the prevalence of
gout varies between 3 and 38 per 1000 in European
and North American populations [1–6 ]. Our study in
Taiwan showed a prevalence of gout of 0.16% in a
rural area, which was significantly diVerent from
suburban (0.67%) and urban ( 0.67%) areas .
The major inhabitants of Taiwan are Taiwanese
whose ancestors migrated from the Fukin Province
400 yr ago. Taiwanese and people (include Hakka)
from mainland China are Han and Mongoloid. The
minority are the aborigines, the original inhabitants of
the island who probably evolved from Austro-Tai
>4000 yr ago . The Taiwan aborigines consist of
nine tribes, each with their own district lifestyle and
language [9, 10]. A survey of a small population [11 ]
indicated a prevalence rate of 94.4% for hyperuricaemia
and 44.4% for gout in Atayal aborigines, one of the
nine tribes, compared to rates reported in Caucasians.
The present study was undertaken to identify
potential risk factors for hyperuricaemia and gout, and to
evaluate the quality of medical care.
MATERIALS AND METHODS
Ho-Ping County in Central Taiwan is composed of
six villages. By 1982, the total population in Ho-Ping
County was 10 149. Among them, 5846 were female
and 4303 were male. Aborigines comprise 32% of the
total population (3253 ) and ~90% of aborigines
belonged to the Atayal tribe. In contrast, the rest
(68%) of the local residents were Han people, consisting
of Taiwanese, Hakka and Chinese who came from
mainland China in 1949.
Aborigines live largely mixed with other races and
random sampling would not be appropriate to study
the epidemiology of disease in this group. However,
>90% of aborigines are Christians and regular church
attendees. In this study, we studied 342 subjects who
were going to church at weekend to maximize
participation of employed individuals, to study a
crosssection of the population and to minimize the costs of
With the help of the health authority in Ho-Ping
County, all subjects were contacted by the local public
health nurse who arranged the evaluation. Health
workers were trained in conducting a standard
interview designed to identify individuals known or
suspected to have gout, and to document risk factors.
Two doctors, three nurses, four laboratory technicians
and two health workers were involved in the study.
The interviewers asked the following questions.
(a) Have you ever had pain or stiVness in any joint or
bone? If yes, where is it? (b ) Have you ever been told
that you had gout or other rheumatic diseases? The
diagnosis of gout was made by identification of
monosodium urate crystals in the synovial fluid, the
presence of tophi or Wallace criteria [12 ]. Further
information included: the clinical presentation (first
involved joint, presence of tophi, disease duration and
medications); family history of relatives with gout; risk
factors (hypertension, hyperlipidaemia, diet, alcohol
and drugs); 15 questions on knowledge about gout.
More than 95% of the aborigines interviewed also
agreed to undergo a physical examination and blood
All subjects gave blood for the determination of uric
acid ( Uricase method, Teco Diagnostics, USA),
triglyceride, cholesterol and creatinine carried out in our
hospital. The uric acid level in normal healthy Chinese
subjects ranges from 3 to 7 mg/ml in men and from 2
© 1998 British Society for Rheumatology
to 6 mg/ml in women before the menopause and from
3 to 7 mg/ml after the menopause. Hyperuricaemia
was defined as uric acid >7 mg/ml in men and
>6 mg/ml in women before the menopause and
>7 mg/ml after the menopause.
Proportions with their 95% confidence interval were
used to estimate the prevalence of gout among
aborigines. Analysis of covariance (ANCOVA) was then
used to compare the diVerences in clinical features
between gout and non-gout, adjusting for the eVect of
age. Finally, logistic regression was carried out to
examine the independent eVect of clinical features on
the prevalence of gout.
The study was carried out between July through
December 1994. There were 342 subjects from four
diVerent aboriginal villages. Among them, 145 (42.4%)
were male and 197 (57.6%) were female ( Table I ).
There were no significant diVerences between males
and females with respect to age distribution (P> 0.05 ).
Gout was confirmed in 40 subjects and most were
male (95%). The diagnosis of gout was based on the
typical clinical presentation. In three cases, urate
crystals were identified in the synovial fluid. Age and sex
distribution in cases with gout is shown in Table I.
There were two peaks in male patients: one is from
the age of 31 to 40 yr (29.0%) and another is from 61
to 70 yr (31.6%). Only two cases with gout were seen
in women and both were post-menopausal.
The mean uric acid level in males was 7.9 ±
1.7 mg/ml, whereas in females it was 5.7± 1.5 mg/ml.
These diVerences (P < 0.001) were significant overall
and in each age group ( Fig. 1), except those over 70 yr
(P = 0.0656). In aborigines, the uric acid started rising
in young adulthood in males. The uric acid was similar
in pre- and post-menopausal women.
The prevalence rate of gout in aborigines was 11.7%
(95% confidence interval 0.08–0.15 ) and was more
common in male aborigines than in females (26.2% vs
1.0%, P < 0.001). Hyperuricaemia occurred in 41.4%
of all aborigines (95% confidence interval 0.36–0.47)
and 53.8% in male aborigines in contrast to 30.7% in
female (P < 0.001).
Among 40 cases with gout, 38 were male and two
were female. The mean age of these patients was 51 yr
Fig. 1.—Serum uric acid level in diVerent age and sex distributions of aborigines. These diVerences (P < 0.001) were significant overall and in
each age group except those over 70 yr (P = 0.0656).
*Statistics for diVerence of age by sex (x2 test, test of homogeneity), P = 0.875.
†P < 0.01 (comparison between male and female with gout).
‡P < 0.05 (comparison between male and female with gout).
Gout (n = 40)
Non-gout (n = 302)
and the mean disease duration was 6 yr. The first
metatarsophalangeal joint was the most frequently
involved during the first attack of gout (53%). Tophi
were seen in 54% of the subjects with gout and were
large and generalized in 70% of patients. The mean
duration from the first gouty attack to the development
of tophi was 4.1 yr. Thirty-five per cent of the
firstdegree relatives in the 40 cases had gout.
Table II lists factors associated with gout. In
univarate analysis, the body mass index (BMI ), triglyceride
and serum cholesterol levels were not diVerent in
aborigines with gout and aborigines without gout.
Factors associated with gout were increased uric acid
and creatinine level, and the percentage of
hypertriglyceridaemia (serum triglyceride >160) and renal
function impairment (serum creatinine >1.5) (P < 0.001
or 0.01 ). In stepwise logistic regression analysis, the
presence of hypertension did not show a significant
diVerence between the two groups. However, for
alcoholism, there was a statistical diVerence between the
two groups (odds ratio 2.07, P < 0.01, 95% confidence
interval 0.49–2.18). Both age and sex also had a
significant diVerence between the two groups (odds
ratio 1.03, 22.83; 95% confidence interval 1.01–1.06,
5.05–103.2; P < 0.05, P < 0.001, respectively). The
survey of dietary habit in the past years showed that,
over the years, people of aborigine and Han descent
have mixed thoroughly and there were no apparent
In 1990, Taiwan had ~330 000 aborigines and 20
million people of other ethnic background. The ancient
aborigines in Taiwan probably moved from the
Southern Provinces of mainland China and then
migrated from Taiwan to the Philippines, Timor, the
Marinas through Micronesia island and New Zealand
. From our previous study and S.-J. Chen’s research
(unpublished ), the high frequency of HLA-A24, B13
and B60 in Taiwan aborigines (including Atayal ),
compared to Taiwanese, suggests that the Taiwan
aborigines are probably more similar to Oceanian
populations than to Taiwanese.
This epidemiological survey was a cross-sectional
study and cases investigated were small numbers of
aborigines living separately in diVerent mountain areas.
It was impossible to carry out random sampling since
the populations have not been fully documented by
census. Therefore, we had to take advantage of when
those aborigines were going to church at the weekend
(they are not working at the weekend ) to obtain
enough samples. In fact, a pilot study was carried out
and we did not find a significant diVerence in dietary
habits (including alcohol ), living standard, education
level, and disease pattern and severity between those
who were going to church and those who were not.
Thus, the data on the number of gouty cases over the
number of investigated cases can represent the
prevalence of gout in this population.
In this study, a high prevalence of hyperuricaemia
was observed in male aborigines. Significantly,
hyperuricaemia in males starts at or before the age of 20 yr
and on average is >2 mg/dl as compared with females
(7.8 vs 5.7). Together with a family history of gout in
35% of first-degree relatives, a genetic defect [13 ], such
as overproduction of uric acid or a renal tubular defect
for uric acid secretion, may be responsible. A study on
115 Maori men in New Zealand by Gibson et al. 
demonstrated that the high percentage of
hyperuricaemia (23%) and gout (8%) in Maori people was
attributable to the lower uric acid clearance. Other reports
[15–22 ] have documented that Polynesians, including
Maoris, Cook Islanders and Samoans, have an
increased frequency of hyperuricaemia. The uric acid
level in South Pacific aborigines ranges from 6.1 to
7.3 mg/100 ml [17 ]. Healey [17 ], Healey and
BayaniSiosan [23 ] and Decker and Lane  reported that
Filipinos in the USA had a higher rate of
hyperuricaemia than Filipinos in the Philippines, and postulated
that it was due to the intake of a more high-purine
Western diet and a defect in the renal excretion of uric
acid. Likewise, Simmonds et al. [25 ] demonstrated that
a high prevalence of hyperuricaemia among Polynesian
women (Maoris, Cook Islanders, Samoans, Tongans)
resulted from a reduced fractional uric acid clearance.
In Taiwan aborigines, three of 342 (0.9%) cases had a
creatinine >1.5. However, when cases with gout were
compared with non-gout, a significant increase in
serum creatinine was also noted in gouty patients [26,
27]. Evaluation of renal function using more sensitive
techniques is needed.
In this study, risk factors for hyperuricaemia, such
as dietary habits, alcohol and hypertension, were
assessed. Eleven (30%) of 37 cases with gout have
hypertension, which is higher than the 24.9% in
aborigines without gout [28, 29]. The prevalence of diabetes
was significantly increased in the aborigines compared
with Han people living in other parts of Taiwan [28,
29 ]. However, there was no diVerence in hypertension
between cases with and without gout. The result is in
contrast to the report by Currie [ 6 ], who observed a
significant increase in hypertension in females with
gout, and RoubenoV  who found that hypertension
was prevalent in gouty patients. In this study, 64% of
gouty patients and 42% of non-gouty aborigines took
at least one half-bottle of high-alcohol liquor or more
than one bottle of beer daily (P < 0.05 ). Although
most aborigines started drinking from the age of 20 yr,
alcohol seemed not to be the single factor in the
development of hyperuricaemia or acute gouty attack
since uric acid was elevated before the age of 20 yr.
The significant increase in hypertriglyceridaemia in
aborigines may be attributable to increased alcohol
intake [31, 32] or unknown factors. A good correlation
between serum uric acid and triglyceride was
demonstrated in aborigines, and is similar to the findings of
Takahashi et al. .
In this study, BMI did not show a significant
diVerence between the aborigines with and without gout.
Obesity is not common in Atayal aborigines, unlike
the Maori who have a high prevalence of
hyperuricaemia and gout and an increased prevalence of obesity
and alcohol consumption [ 14, 22, 34].
The present report confirms that tophi present early
and more severely in aborigines than in Han Chinese.
The percentage of tophi in aborigines with gout was
54%, which is much higher compared to other reports
(17%, 9.2%) [13 ]. The rate of formation of tophaceous
deposits in primary gout is correlated with the degree
and duration of hyperuricaemia. Thus, the prevalent
formation of tophi in aborigines with gout was
attributed to the genetically higher serum uric acid level.
Other possibilities included relatively poor primary
health care for the aborigines. Nearly all the patients
with gout were not being treated for their disease on
a regular basis. We believe that this situation can be
improved after patients frequently receive education
on gout. The other way is to intensify the ability of
local doctors to diagnose and treat gout.
Gout and hyperuricaemia have been recognized as
familial disorders, and the reported frequency of
familial occurrence ranges from 6% to as high as 80% [35 ].
The high prevalence of first-degree relative involvement
in aborigines with gout raises the possibility of an
inborn defect in PRPP synthetase or HGPRT enzyme.
A report by Palmer et al. [36 ] demonstrated that
HGPRTase deficiency was not an important factor in
the development of gout in Maoris. Further research,
except for the enzyme study, including the complete
renal function profile, may confirm the
aetiopathogenesis of hyperuricaemia and gout in Taiwan aborigines.
In summary, gout is severe and prevalent in Taiwan
aborigines. Both genetic and environmental factors are
suggested, and further studies are indicated. The
medical treatment and follow-up appear to be inadequate.
We wish to thank Dr Matthew H. Liang for critically
reviewing the manuscript.
1. National Center for Health Statistics: Damson DA , Adams PF . Current estimates from the National Health Interview Survey , United States , 1986 . Vital and Health Statistics, Series 10 , No. 164. DHHS Publ. No. (PHS ) 87- 1592 . Washington, DC: US Government Printing OYce , 1987 .
2. Lawrence RC , Hochberg MC , Kelsey JL et al. Estimates of the prevalence of selected arthritic and musculoskeletal diseases in the United States . J Rheumatol 1989 ; 16 : 427 - 41 .
3. Kelsey JL , Hochberg MC . Epidemiology and prevention of musculoskeletal disorders . In: Last JM, ed. Public health and preventive medicine . Norwalk , CT: AppletonCentury-Crofts , 1986 : 1227 - 96 .
4. National Center for Health Statistics: Collins JG . Prevalence of selected chronic conditions , United States , 1979 - 1981 . Vital and Health Statistics , Series 10 , No. 155. DHHS Publ. No. (PHS ) 86- 1583 . Washington, DC: US Government Printing OYce , 1986 .
5. Kellgren JH . The epidemiology of rheumatic disease . Ann Rheum Dis 1964 ; 23 : 109 - 22 .
6. Currie JC . Prevalence and incidence of the diagnosis of gout in Great Britain . Ann Rheum Dis 1979 ; 38 : 101 - 6 .
7. Chou CT , Pei L , Chang DM , Lee CF , Schumacher HE , Liang MH . Prevalence of rheumatic diseases in Taiwan: A population study of urban, suburban, rural diVerences . J Rheumatol 1994 ; 21 : 302 - 6 .
8. Bellowood P. The Austronesian dispersal and the origin of languages . Sci Am 1991 ; 256 : 70 - 5 .
9. Ko YC , Liu BH , Hsieh SF . Issues on Aboriginal health in Taiwan . Kaohsing J Med Sci 1994 ; 10 : 337 - 50 .
10. Lin BH , Hsieh SF , Chang SJ , Ko YC . Prevalence of smoking, drinking and betel quid chewing and related factor among Aborigines in Wufeng district . Kaohsing J Med Sci 1994 ; 10 : 405 - 11 .
11. Lai SJ , Ju YJ , Kao JF , Chei TL . The survey of hyperuricemia in Ton-Au Li and Ton-Yei Village of I-Lan city in Taiwan . Yi Chin Bao Taou (Dept Health , Taiwan) 1991 ; 7 : 99 - 105 .
12. Wallace SL , Robinson H , Masi AT , Decker JL , Macarty DJ , Yu TF . Preliminary criteria for the classification of the acute arthritis of primary gout . Arthritis Rheum 1997 ; 20 : 859 - 900 .
13. Kelley WN . Gout and related disorders of purine metabolism . In: Kelley WN, Harris ED , Ruddy S , Sledge CB, eds. Textbook of rheumatology, 1st edn. ( Volume II ) . Philadelphia: WB Saunders Co ., 1981 : 1397 - 436 .
14. Gibson T , Waterworth R , Hatfield P , Robinson G , Bremner K. Hyperuricemia , gout and kidney function in New Zealand Maori men . Br J Rheumatol 1984 ; 23 : 276 - 82 .
15. Rose BS . Gout in Maoris. Semin Arthritis Rheum 1975 ; 5 : 121 - 45 .
16. Lennane GAQ , Rose BS . Gout in the Maori . Ann Rheum Dis 1960 ; 19 : 120 - 5 .
17. Healey LA . Epidemiology of hyperuricemia . Arthritis Rheum 1975 ; 18 (suppl.): 709 - 12 .
18. Prior IAM , Rose BS , Harvey HPB et al. Hyperuricemia , gout, and diabetic abnormality in Polynesian people . Lancet 1966 ; 1 : 33 - 8 .
19. Healey LA , Jones KW . Hyperuricemia in American Samoans . Arthritis Rheum 1971 ; 14 : 283 - 5 .
20. Burch TA , O'Brien WM , Meed R et al. Hyperuricemia and gout in Mariana Islands . Ann Rheum Dis 1966 ; 25 : 114 - 6 .
21. Prior IAM . Epidemiology of rheumatic disorders in the Pacific with particular emphasis on hyperuricemia and gout . Semin Arthritis Rheum 1981 ; 11 : 213 - 9 .
22. Prior IAM , Welby TJ , Ostbye T et al. Migration and gout: The Tokelau Island migrant study . Br Med J 1971 ; 14 : 721 - 6 .
23. Healey LA , Bayani-Sioson PS. A defect in the renal excretion of uric acid in Filipinos . Arthritis Rheum 1971 ; 14 : 721 - 61 .
24. Decker JL , Lane JJ . Gouty arthritis in Filipinos . N Engl J Med 1959 ; 261 : 805 - 6 .
25. Simmonds HA , Mcbride MB , Hatfield PJ et al. Polynesian women are also at risk for hyperuricemia and gout because of a genetic defect in renal tubule handling . Br J Rheumatol 1994 ; 33 : 932 - 7 .
26. Fessel WJ . Renal outcomes of gout and hyperuricemia . Am J Med 1978 ; 67 : 74 - 82 .
27. Yu TF , Berger L. Impaired renal function in gout . Am J Med 1982 ; 72 : 95 - 100 .
28. Miao LJ , Chau JJ , Chang JR . The survey of prevalence of hypertension and diabetes mellitus in 11 Taiwan Aboriginal villages . J Public Health ( Taiwan) 1991 ; 18 : 113 - 31 .
29. Chen CY , Shen YJ . The investigation of health status in Aboriginal villagers of Hwa-Lien city in Taiwan . ChungHwa Chieh Ji 1992 ; 11 : 13 - 9 .
30. RoubenoV R . Gout and hyperuricemia. Rheum Dis Clin North Am 1990 ; 16 : 539 - 51 .
31. Fox IH , John D , Debrune S , Dowosh I , Marliss EB . Hyperuricemia and hypertriglyceridemia: Metabolic basis for the association . Metabolism 1985 ; 34 : 741 - 6 .
32. Gibson T , Graham R. Gout and hyperlipidemia . Ann Rheum Dis 1974 ; 33 : 298 - 303 .
33. Takahashi S , Yamamoto T , Noriwaki Y , Tsutsumi Z , Higashino K. Impaired lipoprotein metabolism in patients with primary gout-influence of alcohol intake and body weight . Br J Rheumatol 1994 ; 33 : 731 - 4 .
34. Faller J , Fox IH . Ethanol-induced hyperuricemia . N Engl J Med 1982 ; 307 : 1598 - 602 .
35. Kelley WN . Inborn errors of purine metabolism . Arthritis Rheum 1977 ; 20 : S221 - 7 .
36. Palmer HL , Fujimoto WY , Palmer DG , Seegmiller JE . Phosphoribosyltransferase levels in Maori subjects with gout . N Engl J Med 1969 ; 70 : 403 - 8 .