вход по аккаунту


Trends in serum uric acid levels.

код для вставкиСкачать
Uric acid levels of adult male volunteers in a
longitudinal study of human aging rose steadily between
1961 and 1W8. In the 1,141 men with 3 serial physical
examinations, who developed no diseases and who took
no drugs known to affect uric acid levels, levels rose
from means below 5.5 mg/dl in 1961-1963 to means
above 6.5 mgldl in 1975-1978. The best predictor of a
longitudinal increase in uric acid level was a gain in
weight, but this, and other significant predictors, explained only a small portion of the increase in this
population. Preliminary data available from a fourth
examination indicate that the rising trend has leveled
Numerous epidemiologic studies have described correlates of serum urate levels in adult men
(1-7). Factors shown to be associated with hyperuricemia include weight, hyperlipidemia, certain diseases
such as atherosclerosis, diabetes, renal diseases, and
essential hypertension, and some drugs, especially
thiazide diuretics. Age, socioeconomk status, and
alcohol consumption have been found by some, but
From the Normative Aging Study, Veterans Administration
Outpatient Clinic, Boston; The Geriatrics Unit of the Massachusetts
General Hospital, Boston; and the Department of Medicine and the
Division on Aging, Harvard Medical School.
Supported by the Medical Research Service of the Veterans
Robert J. Glynn, PhD: Statistician, Normative Aging
Study; Edward W. Campion, MD: Chief, Geriatrics Unit, Massachusetts General Hospital and Instructor of Medicine, Harvard
Medical School; Jeremiah E. Silbert, MD: Director. Normative
Aging Study and Professor of Medicine, Harvard Medical School.
Address reprint requests to Robert J. Glynn, PhD. Normative Aging Study, VA Outpatient Clinic, 17 Court St., Boston. MA
Submitted for publication November 25, 1981; accepted in
revised form August 25, 1982.
Arthritis and Rheumatism, Vol. 26, No. 1 (January 1983)
not all, studies to be positively correlated with uric
acid levels.
Previous epidemiologic studies of uric acid
have been limited by a cross-sectional design which
precludes a determination of factors predictive of
changes in uric acid levels. Furthermore, some studies
have not controlled for the health status and drug
intake of their populations. The current study investigates the uric acid levels of the 2,280 volunteer male
participants in the Normative Aging Study, evaluated
longitudinally between the years 1961 and 1980. Focus
is placed on those 1,141 men with 3 serial examinations (between 3 and 5 years apart) who remained free
of chronic disease and took no drugs that might affect
uric acid levels. In this health and drug screened
group, uric acid levels rose markedly between 1961
and 1978, with population means increasing about I
mg/dl. The aim of this report is to define and attempt to
explain this steady rise in uric acid levels. A further
goal is to describe, using multiple regression techniques, the relative importance of predictors of change
in uric acid levels.
The Normative Aging Study is a longitudinal study of
human aging established by the Veterans Administration in
1961. Six thousand male volunteers from the Greater Boston
area were screened to provide an initially healthy population. Screening was based on laboratory, clinical, radiologic,
and electrocardiographic criteria. Men were disqualified
from participation in the study if they had a history of
conditions such as heart disease, cancer, gout, diabetes,
cirrhosis, peptic ulcer, recurrent asthma, bronchitis, or
sinusitis. Also disqualifying was either systolic blood pressure greater than 140 mm Hg or diastolic blood pressure
greater than 90 mm Hg. Acceptable conditions included
childhood and other generally acute illnesses from which
there were no sequelae. Men were not excluded from
participation in the study because of asymptomatic hyperuricemia.
With these criteria, 2,280 men were accepted into the
Normative Aging Study, ranging in age at entry from 21-81
with a mean of 42 years. Screening and initial examination
occurred between 1961 and 1970. Subsequently, men under
age 52 have repprted for examinations every 5 years; after
age 52 they have reported every 3 years. The educational
level of the population as of 1971 was: 14% less than high
school, 25% high school graduates, 35% beyond high school.
and 26% college graduates. Fifty-two percent of the population were white collar workers (professional. managerial,
clerical, and sales). 'Additional information about the Normative Aging Study design and population has been published elsewhere (8.9).
Serum uric acid levels have been assessed at each
examination in the clinical laboratory of the Veterans Administration Outpatient Clinic by the colorimetric, phosphotungstic acid method. Determinations were performed manually between 1961 and 1970 by the method of Fohn (10). I n
the spring of 1970 the colorimetric method was adapted to
the Technicon Autoanalyzcr ( I I ) (method N-30). Laboratory
quality control procedure consisted of passages of two
known concentrations of human freeze-dried serum with
every 20-30 specimens. Results were reported to the Normative Aging Study only if both standards differed by less
than 2 standard deviations from their known values. New
standards were purchased yearly from the College of American Pathologists which also provided quarterly quality control testing including standard unknowns. Throughout the 20
years of the study the laboratory successfully passed the
quarterly quality' control testing by finding the unknown
standard to differ by less than 2 standard deviations from its
expected value as verified by the College of American
Pathologists. Thus, the uric acid determinations were consistent with those of other laboratories using the same unknowns.
Studies of various uric acid methodologies have
generally found the phosphotungstic acid methods less specific than the uricase procedures (12). In particular, the
phosphotungstic acid method is subject to interference resulting from endogenous nonurate chromagens. nutrients,
and drugs. We compared our phosphotungstic acid method
with the uricase method. Twenty-three blood samples from
23 Nonnative Aging Study volunteers. collected in February
1981, were divided and analyzed by both the colorimetric
method at the Veterans Administration Outpatient Clinic
and the uricase method at the clinical laboratory of the
Massachusetts General Hospital. Uric acid levels were:
uricase method, mean -C SD = 6.20 5 1.31 mg/dl; colorimetric method. mean ? SD = 6.26 ? 1.46. The correlation
between the two methods was 0.93 and the 95% confidence
interval for the mean difference between the two methods.
based on paired r-test, was (-0.160, 0.273). Thus, the two
methods were quite comparable.
Although the Normative Aging Study population was
screened at entry, many men developed diseases between
their first and third examinations. To ensure that changes in
uric acid levels were independent of disease and medication.
the population was further screened. The 1.5 17 men who had
3 complete examinations at regular intervals were considered. Between their first and third examinations, 376 of these
men either developed one of the following conditions: diabetes, hypertension, psoriasis, gout, ischemic heart disease.
arthritis or kidney disease: or were taking certain medication: allopurinol, probenecid, o r antihypertensive drugs including thiazides. Excluding these participants. the screened
population consisted of 1,141 men with 3 serial examinations.
Variables examined as possible predictors of change
in uric acid were: I ) age at first examination, 2) average (over
3 examinations) fasting serum triglyceride level, 3) average
(over 3 examinations) blood urea nitrogen (BUN) level, 4)
body mass index (weight divided by the square of height) at
first examination, 5 ) change in body mass index between
first and second examinations, 6) alcoholic beverage consumption, 7) cigarette consumption, and 8 ) socioeconomic
Fasting serum triglyceride levels have been assessed
at each examination since 1%5 and BUN levels have been
assessed since 1968. Because levels of these two variables
were not available on all participants for first examination,
and additionally to control for the variability of these measures, the average of a n individual's availablc measures on
each variable was considered as a predictor.
T w o weight change variables were used: "increase in
body mass index" was defined for each individual as the
increase in his body mass index if he gained weight, and zero
if he lost weight; "decrease in body mass index" was
defined for each individual a s his decrease in body mass
index if he lost weight and Lero if he gained weight.
Alcoholic beverage consumption was assessed by
questionnaire in June 1973. Participants described the usual
number of drinks of beer, wine, and spirits consumed per
day, week, month. or year. A composite scale for all
beverages, measured in drinkslday, was formed for each
individual. Cigarette consumption was assessed in packslday
at the time of the second examination. Both the drinking and
smoking variables were transformed by the natural logarithm
to normalize their distributions. Quitting smoking between
the first 2 examinations (scored by means of a yeslno
indicator variable) was considered an independent predictor.
The Duncan index was used to measure socioeconomic status (13). Scores on this index range from I to 100.
with high scores corresponding t o high status jobs.
Regression analysis was used to analyze the association between uric acid level and test year. Multiple regression was used to relate predictors to individuals' changes in
uric acid levels (14). Uric acid level at first examination was
entered as a predictor to control for regression to the mean
by the method of Rosner (15). All analyses were performed
using the General Linear Models procedure in the SAS
statistical package (16).
Mean uric acid levels b y e x a m i n a t i o n y e a r for
all uric acid levels assessed during t h e first 3 N o r m a tive Aging S t u d y e x a m i n a t i o n s are shown in Figure l .
There were a total of 5,941 uric acid determinations on
the 2,280 men initially admitted to the study. This
includes data from men having fewer than 3 examinations as well as men who had all their examinations. It
also includes data from men who developed chronic
diseases or took drugs affecting uric acid levels sometime before their third examination. With the exception of 1964 when no examinations were performed,
there were over 100 uric acid determinations for each
year between 1962 and 1979.
Mean uric acid levels rose steadily and substantially from 1961 until 1976. Mean levels in 1961-1963
were about 5.5 mg/dl, whereas in 1975-1978 they were
above 6.5 mg/dl. Also shown in Figure 1 is the
regression line of uric acid levels over time. The
correlation between test year and uric acid level is 0.31
and the associated P value is less than 0.001, indicating
a strong positive relation between levels and time. The
line fits the mean levels well, except at the extreme
years when there were relatively few examinations.
Since diseases and medications affect uric acid
levels, we examined the levels for those participants
who had all 3 examinations and who maintained the
criteria for health and absence of medication described
above. Mean uric acid levels by examination year for
this screened group of 1,141 men are shown in Figure
2. In this group also, the same steady upward trend in
levels is apparent. Again, mean levels in 1961-1963
were below 5.5 mg/dl, whereas in 1975-1978 they were
above 6.5 mg/dl. Thus, the increase could not be
accounted for by known health decline or medication.
6 8-
~ ~ ~ " ' " " ' ' ' '
61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 Bo
161 (69 (441
Figure 2. Mean uric acid levels by examination year from first
three Normative Aging Study examinations for screened population
of 1,141 men; 3,423 determinations.
The slope of the regression line of uric acid levels on
time was still significantly different from zero ( P <
0.001) and the correlation between levels and test year
was 0.30. When a weighted regression was used to
control for greater variability of values at the extreme
years, the estimated regression slope and intercept
were nearly identical. Uric acid means and standard
deviations by specific examination for the screened
group were: exam 1 , mean ? SD = 5.77 ? 0.87 mg/dl;
exam 2, mean -+ SD = 6.21 ? 1.03 mg/dl; exam 3,
mean 5 SD = 6.53 ? 1.15 mg/dl. Exam 1 took place
between 1961 and 1970 with subsequent examinations
every 3-5 years.
70To investigate whether trends in uric acid levels
68 might be due to a cohort effect, we considered mean
uric acid levels for each year within four birth cohorts.
Plots of these means are shown in Figure 3 . Means are
$, 6 4 plotted for a specified cohort in a given year only if
.as 6 2 there were at least five uric acid determinations in the
pb 6 0 year for that cohort. Some of the fluctuation in the
g 58curves is due to small numbers of determinations for
56 some cohorts in certain years. This is particularly true
of the oldest cohort. In spite of some small sample
sizes, there is a clear upward trend in each cohort. The
, , , , , , ,
, ,
of the curves indicates that the trend of
increasing uric acid levels is independent of cohort
YEAR 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
(22l(1841(KHI ~ Y l l ~ 7 l 1 ~ ~ 6 5 a ( 5 6 1 ~ ~ 1 9 3 ~ L 2 2 4 1 1 4 ( 1 ~ 5 2 6 ) ( 5 ( 2 1 ( 3 ~ ~ l Z l l ~ Z Z l effects.
Multiple regression was used to determine predictors of changes in uric acid levels between exams 1
Figure 1. Mean uric acid levels by examination year from first three
Normative Aging Study examinations: 5,491 determinations.
and 3 in the screened group. Results of this analysis
L G!
7 0 7 4
Figure 3. Mean uric acid levels by examination year within four
birth cohorts from first three Normative Aging Study examinations
for screened population of 1,141 men. Number of determinations per
year is indicated.
are shown in Table 1. Exam 3 uric acid was the
outcome or dependent variable. It was hypothesized
that a weight gain might have a different marginal
effect on changes in uric acid than a weight loss, so
separate weight gain and weight loss variables were
considered. By far the best predictor of a high exam 3
uric acid level, controlling for exam 1 level, was an
increase in body mass index. Other significant predictors were the number of alcoholic drinks per day, the
average serum triglyceride level, and initial body mass
index. With these variables in the model, BUN, decrease in body mass index, cigarette consumption,
quitting smoking, socioeconomic status, and age were
not significantly associated with exam 3 uric acid
These variables could explain only a small
portion of the rise in uric acid levels. The R-square of
the model was 0.296 and the variables other than exam
1 uric acid explained only 9.7% of the variance unexplained by exam 1 uric acid. A separate regression of
uric acid level on test year was performed which
included only those men in the screened population
whose weight changed less than 1 kg/m2 and who
drank less than one drink per day. Even in this group
there were substantial rises in uric acid levels over
In Table 2 are shown descriptive statistics of
the variables found in the above regression to be
significantly associated with changes in uric acid levels. Means and standard deviations of important predictors are shown within four categories of initial uric
acid level and three categories of change in uric acid
level (exam 3 level-exam I level). This stratification
was done to control for the association between
change and initial level. Variables with notable differences between groups are change in body mass index
and alcoholic drinkdday. Within each of the four
initial uric acid categories, those men whose uric acid
levels rose more than 1 mg/dl had mean body mass
index changes between 0.31 and 0.66 kg/m2 greater
than those men whose uric acid levels declined more
than 0.5 mg/dl. Similarly, within each initial level
category, those men whose uric acid levels rose more
than 1 mg/dl drank between 0.11 and 1.35 drinkdday
more than those men whose uric acid levels declined.
Men whose uric acid levels rose also had, on average,
higher serum triglyceride levels and higher initial body
mass indexes than men whose uric acid levels declined.
Although these trends were consistent, there
was substantial within-category variance for each of
these four variables. For example, within each initial
uric acid category, there were men whose body mass
index declined at least I kg/m’ yet their uric acid level
rose more than 1 mg/dl. This gives further evidence
that, although on average, these variables are predictive of average changes in uric acid levels, they are not
Table 1. Multiple regressjon predicting uric acid level for exam 3
Exam I uric acid
Serum triglyceride
Body mass index (kg/mz)
Exam 1
Alcoholic drinkdday
(transformed by log)
Cigarette packdday
(transformed by log)
Indicator of quitting smoking
Socioeconomic status
(Duncan index)
error of
Table 2. Variables associated with changes in uric acid levels
Uric acid levels (mg/dl)
First examination
4 . 0
Exam 3 - exam I
-0.5-1 .O
-0.5-1 .O
Body mass index
(mean t SD)
t 3.77
? 2.74
2 2.11
2 2.58
t 2.77
t 3.20
2 2.47
2 2.21
t 2.24
? 2.98
2 3.06
sufficient to account for the uric acid level increases in
this population.
In the plots of mean uric acid levels over time
using data from the first three Normative Aging Study
examinations shown in Figures 1 and 2, the means in
1979 and 1980 hint that the increasing trend in uric acid
levels has stopped. Most Normative Aging Study
participants had their third examination before 1979;
thus, there were relatively few examinations in 1979
and 1980. To further investigate the association between uric acid levels and time, we considered the
available data from participants who had a fourth
examination. As of February 1981. almost half of the
Normative Aging Study participants have had a fourth
examination. In Figure 4 are shown mean uric acid
levels by test year from the first 4 examinations of all
Normative Aging Study participants. The data indicate
that the increasing trend has indeed leveled off. In fact,
means have declined slightly since the peak years of
1974-1978. Since a line no longer adequately describes
the relation between uric acid levels and test year, a
quadratic curve has been fit to the data by least
squares. Both the linear and quadratic terms of this
curve are significantly different from zero (each P <
0.001) indicating that uric acid levels are varying over
time. The four examination data from the screened
population were similar and are not shown.
This study describes and attempts to explain
the puzzling observation that uric acid levels of male
volunteers in the Normative Aging Study rose steadily
and substantially between the years 1961 and 1978.
The increase could not be explained by the development of disease in this initially health-screened popula-
Change in body
mass index
(mean 2 SD)
Serum triglyceride
(mean t- SD)
(mean t SD)
0.24 2 1.27
0.31 f 1.27
0.55 2 1.43
0.33 ? 1.10
0.28 2 1.19
0.71 t 1.51
-0.02 2 1.34
0.23 ? 1.32
0.63 2 1.64
-0.08 2 1.30
0.29 2 1.15
0.58 t 1.35
102.1 t 11.4
135.4 t- 44.8
134.8 t 46.5
129.8 5 77.9
127.0 t- 39.6
145.3 3- 51.5
142.0 t 50.8
143.8 t 51.6
151.5 t 64.6
145.3 2 44.7
152.9 2 50.4
159.7 t 57.3
0.52 t 0.63
1.12 t 1.43
1.00 t 1.14
0.97 2 1.43
1.41 ? 1.75
1.04 2 1.14
1.18 -C 1.37
1.15 t 1.19
1.40 2 1.33
1.31 t 1.26
2.75 t 4.22
tion since the same pattern and amount of increase
appeared in those men who developed no chronic
disease and who took no drugs affecting uric acid
levels during the period of observation. When separate
birth cohorts were considered, similar trends in uric
acid levels over time appeared in each cohort. Uric
acid data from 1979 and 1980 indicate that the increase
in mean levels has peaked in 1976-1978 and is now
The best predictors of an increase in uric acid
level between exams I and 3 were an increase in body
mass index between exams 1 and 2, alcohol consumption level, triglyceride level, and initial body mass
index. These results extend a longitudinal framework
to the cross-sectional findings of other studies (1-7).
61 62 63 64 65 66 67 68 69 70 71 72
74 75
77 78 79 80 81
(3301~11116501661)1 1 ~ ~ ~ 1 1 ~ 5 2 6 ~ 1 5 a X 3 4 2 X 2 8 5 ~ 1 4 1 6 )1391
Figure 4. Mean uric acid levels by examination year from first four
Normative Aging Study examinations; 6,879 determinations.
Noteworthy is the examination of the relationship
between weight changes and uric acid changes, not
testable in previous cross-sectional studies. In contrast to the importance of weight gain for uric acid
increase, weight loss was not significantly related to
decreased exam 3 uric acid level, indicating that uric
acid levels may be more affected by weight gain than
Potential causes of this population’s increasing
unc acid levels could include subclinical renal deterioration, laboratory artifact, dietary changes, and environmental factors. A gradual, progressive deterioration of renal function with normal aging (17) has been
well documented. Messerli et a1 (18) described an
association between mild asymptomatic hyperuricemia and decreased renal blood flow. However, it is
difficult to hypothesize that this is a cause of rising uric
acid levels in a healthy population, particularly in the
absence of an association between uric acid levels and
age. Other investigators are in agreement with the
present study in finding no association between normal aging and uric acid levels (except for the increase
in females after menopause.)
An increase resulting from laboratory artifact
appears unlikely since it would be expected that an
artifactual trend would display a discontinuous curve
over time instead of the smooth rise actually observed.
There are diurnal variations in uric acid levels (19), but
throughout the course of the Normative Aging Study,
blood samples have been drawn between 7:45 and 8: 15
A M . Bywaters and Holloway (20) conducted a multicenter assessment of uric acid measurements and
found that the method of Fohn adapted to an autoanalyzer (the method used by our study since 1970)
showed the greatest precision and reproducibility between centers. They did, however, conclude that a
uricase technique is preferable for special projects
where absolute uric acid values are required. Split
samples comparing results from our method with a
uricase procedure showed excellent agreement (r =
The involvement of dietary factors in an increase in uric acid levels is suggested by the importance of weight gain and alcohol consumption in
predicting uric acid increases. However, weight gain
and alcohol consumption alone accounted for only a
small percentage of the increase in mean levels of uric
acid. More detailed dietary information might have
been more predictive but was beyond the scope of this
study. In a cross-sectional study of correlates of serum
uric acid levels among 8,000 Japanese-American men
in Hawaii, Yano et a1 (21) examined a 24-hour dietary
history from each participant. They found relative
weight, alcohol consumption level, and serum triglyceride level to be the variables most strongly associated
with uric acid level. After controlling for these variables, no measure of current diet, including total calories, proportion of saturated fat, and proportion of
animal protein, was found to correlate significantly
with uric acid level.
There have been some suggestions that environmental factors may influence serum uric acid levels. Goldstein et a1 (22) examined serum urate measurements performed at biweekly intervals throughout
a 12-month period in 12 adult men. They found uric
acid levels very labile with a significant seasonal effect
showing the highest levels in the summer. Fessel et a1
(23) have presented data on over 30,000 uric acid
determinations on men in San Francisco and Oakland
examined between 1964 and 1968. Significant differences were observed both over time and between
cities. Mean levels in San Francisco vaned by as much
as 0.96 mg/dl in different years. In Oakland, with more
than 4,000 observations in each year, mean uric acid
levels were fairly constant between 1964 and 1967, but
rose 0.24 mg/dl, from 5.78 mg/dl to 6.02 mg/dl, between 1967 and 1968. There were also differences in
mean levels between the two cities as large as 0.62
mg/dl in the same year. This emphasizes the need for
further exploration of the effect of local environmental
factors on uric acid levels.
The phenomenon of longitudinally rising levels
of serum uric acid results in an increasing number of
persons entering a range in which clinicians may be
tempted to initiate hypouricemic therapy. In our
health-screened cohort, by the time of the third examination, 11.3% of men had serum uric acid levels of 8.0
mg% or greater and 2.3% of men had levels of 9.08mg%
or greater, compared with only 1.2% and 0.3%, respectively, at the initial examination. This is additional
reason for clinicians to arrive at a clear policy concerning management of asymptomatic hyperuricemia. Current opinion generally favors no treatment for asymptomatic hyperuricemia (24).
The trend of increasing uric acid levels among
Normative Aging Study participants is an example of a
period effect, i.e., a change occurring in a given time
interval which is parallel in several age groups. Period
effects often occur artifactually and so must be validated carefully. Cooper (25) writes about another currently occurring period effect, namely the changing rates
of heart disease in the U.S. (26) and the Soviet Union.
He suggests that studying period effects is particularly
important because they provide invaluable clues to
causation a n d thereby suggest preventive measures.
1. Myers AR, Epstein FH, Dodge JH, Mikkelsen WM: The
relationship of serum uric acid to risk factors in coronary
heart disease. Am J Med 45520-528, 1968
2. Acheson RM, Florey C du V: Body weight, ABO blood
groups, and altitude of domicile as determinants of
serum-uric-acid in military recruits in four countries.
Lancet 2:391-395, 1969
3. Hall AP, Barry PE, Dawber TR, McNamara PM: Epidemiology of gout and hyperuricemia: a long-term population study. Am J Med 42:27-37, 1967
4. Munan L, Kelly A, Petitclerc C: Population serum urate
levels and their correlates: the Sherbrooke regional
study. Am J Epidemiol 103:369-382, 1976
5. Fessel WJ, Barr GD: Uric acid, lean body weight, and
creatinine interactions: results from regression analysis
of 78 variables. Semin Arthritis Rheum 7: 115-121, 1977
6. Kelley WN, Weiner IM, editors: Uric Acid. Berlin,
Springer-Verlag, 1978
7. Scott JT: Long-term management of gout and hyperuricemia. Br Med J 281:1164-1166, 1980
8. Bell B, Rose CL, Damon A: The Normative Aging
Study: an interdisciplinary and longitudinal study of
health and aging. Aging Hum Devel 3:5-17, 1972
9. Bosst R, Ekerdt DJ, Silbert JE: The Veterans Administration Normative Aging Study, Longitudinal Research
in the United States. Edited by A Mednick, M Harway.
Martinus Nijhoff (in press)
10. Folin 0: Standardized method for determination of uric
acid in unlaked blood and in urine. J Biol Chem 101:111125, 1933
11. Hawk PB, Oser BL, Summerson WHG: Practical Physiological Chemistry. Thirteenth edition. Philadelphia,
Blakiston, 1954, p 564
12. Geisinger KR. Batsakis JG, Bauer RC: Serum uric acid.
Am J Clin Pathol 72:330-336, 1979
13. Reiss A, Duncan 0, Hatt P, North C: Occupations and
Social Status. New York, Free Press of Glencoe, 1961,
pp 109-138
14. Snedecor GW, Cochran WG: Statistical Methods. Sixth
edition. Ames, Iowa State University, 1967, pp 381-418
15. Rosner B: The analysis of longitudinal data in epidemiologic studies. J Chron Dis 32:163-173, 1979
16. SAS Institute Inc: SAS User’s Guide, 1979 Edition.
Cary, NC, SAS Institute Inc, 1979, pp 237-263
17. Rowe JW, Andres R, Tobin JD, Norris AH, Shock NW:
The effect of age on creatinine clearance in men: a crosssectional and longitudinal study. J Gerontol 31:155-163,
18. Messerli FH, Frohlich ED, Dreslinki GR, Suarez DH,
Aristimundo GG: Serum uric acid in essential hypertension: an indicator of renal vascular involvement. Ann
Intern Med 93:817-821, 1980
19. Rubin RT, Plag JA, Arthur RJ, Clark BR, Rahe RH:
Serum uric acid levels: diurnal and hebdomadal variability in normoactive subjects. JAMA 208:1184-1186, 1969
20. Bywaters EGL, Holloway VP: Measurement of serum
uric acid in Great Britain in 1963. Ann Rheum Dis
23:236-239, 1964
21. Yano K, Rhoads GG, Kagan A: Epidemiology of serum
uric acid among 8000 Japanese-American men in Hawaii. J Chronic Dis 30:171-184, 1977
22. Goldstein RA, Becker KL, Moore CF: Serum urate in
healthy men: intermittent elevations and seasonal eEect.
N Engl J Med 287:649-650, 1972
23. Fessel WJ, Siegelaub AB, Johnson ES: Correlates and
consequences of asymptomatic hyperuricemia. Arch
Intern Med 132:44-54, 1973
24. Liang MA, Fries JF: Asymptomatic hyperuricemia: the
case for conservative management. Ann Intern Med
88:666-670, 1978
25. Cooper RC: Rising death rates in the Soviet Union. N
Engl J Med 304:1259-1265, 1981
26. Stern MP: The recent decline in ischemic heart disease
mortality. Ann Intern Med 91:630-640, 1979
Без категории
Размер файла
638 Кб
acid, trends, uric, level, serum
Пожаловаться на содержимое документа