 |
|
Fasting and 2-Hour Postchallenge Serum Glucose Measures and Risk of Incident Cardiovascular Events in the Elderly
The Cardiovascular Health Study
Nicholas L. Smith, PhD, MPH;
Joshua I. Barzilay, MD;
Douglas Shaffer, MD, MHS;
Peter J. Savage, MD;
Susan R. Heckbert, MD, PhD;
Lewis H. Kuller, MD, DrPH;
Richard A. Kronmal, PhD;
Helaine E. Resnick, PhD, MPH;
Bruce M. Psaty, MD, PhD
Arch Intern Med. 2002;162:209-216.
ABSTRACT
| |
Background The contributions of fasting and 2-hour postchallenge glucose level
to cardiovascular events remain ill-defined, especially for nondiabetic adults.
This study examined the relative predictive power of fasting and 2-hour glucose
level on cardiovascular event risk.
Methods A total of 4014 community-dwelling adults 65 years or older who participated
in the baseline visit of the Cardiovascular Health Study and who were without
treated diabetes or previous myocardial infarction or stroke were eligible
for analyses. Participants with treated diabetes at baseline were excluded.
Incident myocardial infarction or stroke, or coronary death, was the outcome
of interest. Age-, sex-, and race-adjusted proportional hazards regression
models described individual and joint associations between baseline measures
of fasting and 2-hour postchallenge glucose level and event risk.
Results There were 764 incident cardiovascular events during 8.5 years of follow-up.
Fasting glucose level of 115 mg/dL (6.4 mmol/L) or more was associated with
an increased cardiovascular risk (hazard ratio [HR], 1.66 [95% confidence
interval (CI), 1.39-1.98]) in adjusted analyses compared with fasting glucose
level less than 115 mg/dL. Two-hour glucose level was associated with a linear
risk (HR, 1.02 [95% CI, 1.00-1.04] per 10 mg/dL [0.6 mmol/L]) that included
an additional increase in risk for 2-hour glucose level of 154 mg/dL (8.5
mmol/L) or more (HR, 1.29 [95% CI, 1.04-1.59]) in adjusted analyses. In joint
fasting and 2-hour glucose models, only 2-hour glucose level remained predictive
of event risk.
Conclusions Two-hour glucose level was better than fasting glucose level alone at
identifying older adults at increased risk of major incident cardiovascular
events.
INTRODUCTION
IN 1997, THE American Diabetes Association (ADA) revised diabetes diagnostic
criteria by lowering the fasting plasma glucose threshold and by recommending
against the routine use of glucose challenge testing, especially as it concerns
epidemiologic studies.1 The change was justified
primarily by epidemiologic data indicating that the new criteria would identify
roughly the same number of people with microvascular complications as the
previous criteria, without the burden of glucose challenge testing. These
revised criteria have prompted the research community to examine the effects
of the diagnostic change on the identification of persons at risk for cardiovascular
disease (CVD), the primary complication of glucose disorders in old age, and
all-cause mortality.2-3 Findings
indicate that the current ADA criteria do not maximize the identification
of persons at risk for these major health outcomes. Nonetheless, the individual
and joint associations of fasting and 2-hour glucose measures with cardiovascular
morbidity and mortality remain ill-defined, especially for those who do not
meet 1997 ADA diabetes criteria.
In this study, we examined the relative predictive power of fasting
and 2-hour postchallenge glucose level on fatal and nonfatal myocardial infarction
and stroke and on cardiovascular mortality among a cohort of older adults,
none of whom had treated diabetes at study entry.
SUBJECTS AND METHODS
SETTING AND DESIGN
The Cardiovascular Health Study (CHS) is a population-based, prospective
cohort study of risk factors for CVD in the elderly.4
Participants were recruited from 4 US communities (Washington County, Maryland;
Pittsburgh [Allegheny County], Pa; Forsyth County, North Carolina; and Sacramento
County, California) on the basis of a randomly generated sampling frame from
Health Care Financing Administration files. Annual examinations began in June
1989 and ended in June 1999.
SUBJECTS
The CHS cohort consisted of 5201 community-dwelling adults 65 years
and older who participated in the baseline clinic visit in 1989 to 1990 (original
cohort) and an additional 687 African American adults 65 years or older who
were recruited in 1992 to 1993 (new cohort). Baseline postchallenge glucose
testing was not done in the new cohort, so these participants were not included
in this report. Approximately 57% of eligible participants participated in
the study. Nonparticipants were more likely to be older and less educated
and to have more self-reported CVD and hypertension.5
All study participants gave informed consent for their participation according
to guidelines of the appropriate institutional review boards.
This study excluded participants with missing fasting or 2-hour glucose
measures (see "Measures") and those who reported using insulin or an oral
hypoglycemic agent at the baseline medication inventory.6
Self-reported history of diabetes was not a criterion for exclusion, although
28% of participants with self-reported diabetes and not using antidiabetic
medication opted not to undergo glucose challenge testing. The study population
was further restricted to participants who did not have a baseline history
of myocardial infarction or stroke, to investigate the risk of incident cardiovascular
events.7 We did not exclude participants with
a history of other clinical or subclinical cardiovascular conditions, since
we were interested in the primary prevention of major coronary and cerebrovascular
atherosclerotic or thrombotic events, namely, myocardial infarction and stroke.
MEASURES
At the baseline examination, venipuncture was performed on study participants
under 12-hour fasting conditions early during the study visit. Serum glucose
was measured (Kodak Ektachem 700 Analyzer; Eastman Kodak Corp, Rochester,
NY).8 After the fasting venipuncture, 75 g
of glucose was given orally to consenting nondiabetic participants. A second
venipuncture was performed 2 hours after the glucose challenge.
Covariate baseline measures included demographic characteristics (age,
sex, race [white vs other], and self-reported health [good, very good, or
excellent vs fair or poor]), cardiovascular risk factors (body mass index
[weight in kilograms divided by the square of height in meters], current smoker,
currently treated hypertension, sitting diastolic and systolic blood pressure,
and low-density lipoprotein cholesterol level), clinical CVD (history of angina,
coronary revascularization, congestive heart failure, or transient ischemic
attacks; electrocardiogram-identified atrial fibrillation), and subclinical
CVD (no clinical CVD, ankle-arm index  0.9, maximum internal or common
carotid wall thickness >80th percentile [ 2.06 and 1.16 mm, respectively],
carotid stenosis >25%, major electrocardiogram abnormalities [ventricular
conduction defect, major Q-wave abnormalities, left ventricular hypertrophy,
isolated ST-T–wave abnormalities, atrial fibrillation, and atrioventricular
block], abnormal ejection fraction or wall motion on echocardiogram [qualitatively
assessed], or positive findings on Rose Questionnaire for angina or claudication9). These covariates were selected to adjust for potential
confounding of the glycemia–CVD event risk association in selected analyses.
An incident coronary artery disease (CAD) event was defined as first
nonfatal or fatal myocardial infarction or CAD death; an incident cerebrovascular
disease (CBD) event was defined as first nonfatal or fatal stroke of any type.10 A combined end point for an incident cardiovascular
event was based on the first occurrence of either CAD or CBD.4
Data on adjudicated events through June 1998 are presented in this report.
These analyses were based on the updated CHS database as of June 1999, which
incorporated minor corrections to baseline examination data.
ANALYSIS
Descriptive summaries of participant characteristics were derived for
the 3 categories of fasting glucose level described in the revised ADA diabetes
criteria1: normal, 109 mg/dL (6.0 mmol/L) or
less; impaired fasting glucose, 110 to 125 mg/dL (6.1-6.9 mmol/L); and diabetes,
126 mg/dL (7.0 mmol/L) or more.
Unadjusted cardiovascular event rates per 1000 person-years were calculated
independently for each decile of fasting and 2-hour glucose level. Exploratory
statistical testing compared threshold (single step), linear, linear plus
quadratic, and linear plus threshold (linear model with jump in risk at specified
cutoff point) models to determine which model best fit the data. For proportional
hazards modeling, failure time was the date of the incident fatal or nonfatal
CVD event. All surviving participants were censored on June 30, 1998, the
last day of follow-up for these analyses, or on the date of their last contact
with the study if they were unavailable for follow-up. Survival time was the
difference in days between date of enrollment and either failure or censoring
date.
Multivariate proportional hazards regression was used to determine the
individual and joint contributions of fasting and 2-hour glucose levels to
cardiovascular risk. Quintile and continuous measures of glucose were used,
and hazard ratios (HRs) and their corresponding 95% confidence intervals (CIs)
were calculated for each quintile. Models were adjusted for age, sex, and
race and were further adjusted for potential confounders to determine their
effect on the glycemia-CVD relationship.
Unadjusted CVD event rates are presented across strata of fasting glucose
level according to 2-hour glucose level categories recommended by World Health
Organization criteria: normal, 139 mg/dL (7.7 mmol/L) or less; impaired glucose
tolerance, 140 to 199 mg/dL (7.8-11.0 mmol/L); and diabetic, 200 mg/dL (11.1
mmol/L) or more. Age-, sex-, and race-adjusted hazard ratios for 2-hour glucose
level categories were estimated by means of proportional hazards regression.
Population attributable risk percentages were used to estimate the percentage
of incident events that could be explained by impaired glucose tolerance or
diabetic-level, 2-hour glucose concentration among participants with normal
fasting and impaired fasting levels.11 Percentages
were calculated for each 2-hour glucose level category on the basis of estimated
hazard ratios from models that adjusted for potential confounders listed above.
Isolated postchallenge hyperglycemia was defined as fasting glucose
measures of 125 mg/dL (6.9 mmol/L) or less and 2-hour glucose measures of
200 mg/dL (11.1 mmol/L) or more. Age-, sex-, and race-adjusted hazard ratios
were calculated for isolated postchallenge hyperglycemia compared with 2-hour
glucose values of 139 mg/dL (7.7 mmol/L) or less and fasting glucose values
of 125 mg/dL (6.9 mmol/L) or less.
RESULTS
There were 386 participants (7% of the original cohort) with pharmacologically
treated diabetes at baseline and 585 (11% of the original cohort) with a baseline
history of myocardial infarction or stroke who were not eligible for these
analyses. From among the 4230 remaining participants, 184 participants (4%)
who were missing fasting or 2-hour glucose measures and 32 (1%) who were fasting
less than 8 hours were excluded. These exclusions yielded 4014 participants
for analysis (95% of eligible participants). During a median follow-up of
8.5 years, there were 359 incident myocardial infarctions (9%), of which 12%
were fatal; 348 incident CBD events (9%), of which 12% were fatal; and 117
CAD deaths (3%). There were 60 participants who had both incident CAD and
CBD events during follow-up. Thus, there were a total of 764 incident CVD
events (19%), of which 27% were fatal, among the 4014 study participants.
The date of the first occurring event was used for combined end points.
Table 1 depicts characteristics
of the study population according to fasting glucose values. As expected,
CVD risk factors including male sex, hypertension, higher systolic blood pressure
and body mass index, and lower high-density lipoprotein cholesterol levels
were generally more prevalent across categories of increasing fasting glucose
levels. Smoking and low-density lipoprotein cholesterol risk factors were
not associated with fasting glucose levels. The prevalence of clinical CVD
did not differ according to baseline glucose measures, whereas subclinical
CVD in the absence of clinical disease was increasingly more common among
those with higher fasting glucose level. Noteworthy is that more than half
of the cohort had clinical or subclinical CVD at study entry.
|
|
|
|
Table 1. Baseline Characteristics, Cardiovascular Health Study, 1989
to 1990*
|
|
|
INDEPENDENT CONTRIBUTIONS OF FASTING AND 2-HOUR GLUCOSE LEVELS
Figure 1 presents the unadjusted
incident CVD event rates per 1000 person-years for each decile of fasting
and 2-hour glucose levels independently. For fasting and 2-hour glucose measures,
the event rate appeared to increase rapidly in the top 2 and 4 deciles, respectively.
Nearly identical associations were seen for CHD-specific and CBD-specific
rates (Figure 2) and sex-specific
rates (Figure 3). When data were
stratified by prevalent CVD (clinical and subclinical), the data depicted
a similar increase in event rates in the upper deciles, but only among subjects
with CVD; among those without CVD, rates across glucose deciles were variable
but did not suggest a trend (Figure 4).
|
|
|
|
Figure 1. Cardiovascular event rates per
1000 person-years according to deciles of fasting and 2-hour glucose level,
Cardiovascular Health Study, 1989 to 1998. To convert glucose level to millimoles
per liter, multiply by 0.0555.
|
|
|
|
|
|
|
Figure 2. Coronary (CHD) and cerebrovascular
(CBD) event rates per 1000 person-years according to deciles of fasting and
2-hour glucose level, Cardiovascular Health Study, 1989 to 1998. To convert
glucose level to millimoles per liter, multiply by 0.0555.
|
|
|
|
|
|
|
Figure 3. Cardiovascular event rates per
1000 person-years according to deciles of fasting and 2-hour glucose level
stratified by sex, Cardiovascular Health Study, 1989 to 1998. To convert glucose
level to millimoles per liter, multiply by 0.0555.
|
|
|
|
|
|
|
Figure 4. Cardiovascular event rates per
1000 person-years according to deciles of fasting and 2-hour glucose level
stratified by the prevalence of clinical or subclinical cardiovascular disease
(CVD), Cardiovascular Health Study, 1989 to 1998. To convert glucose level
to millimoles per liter, multiply by 0.0555.
|
|
|
Exploratory analyses of age-, sex-, and race-adjusted fasting glucose
data demonstrated that a simple threshold model where the risk of an incident
CVD event increased at the 85th percentile of fasting glucose level (115
mg/dL [6.4 mmol/L]) was the best fitting model. This stepped increase
was associated with a 66% increase in CVD event risk (HR, 1.66; 95% CI, 1.39-1.98)
compared with those with baseline fasting glucose level less than 115 mg/dL.
For CAD and CBD outcomes, this corresponded to a 62% and 84% increase in risk,
respectively. No interactions were detected between a fasting glucose threshold
and either sex or CVD for all cardiovascular events.
For 2-hour glucose level, age-, sex-, and race-adjusted analyses demonstrated
that a combined linear and threshold model fit the data best. In this model,
there was a 2% increase in risk of an incident CVD event for every 10-mg/dL
(0.6-mmol/L) increase in 2-hour glucose level (HR, 1.02; 95% CI, 1.00-1.04)
plus an additional 29% increase in risk for all values above the 65th percentile
(154 mg/dL [8.5 mmol/L]) (HR, 1.29; 95% CI, 1.04-1.59). This held
true for CAD and CBD outcomes where there was a 2% linear increase in addition
to a 28% (CAD) and 42% (CBD) increase in event risk for 2-hour glucose values
of 154 mg/dL (8.5 mmol/L) or more. No interactions were detected between linear
and threshold 2-hour glucose levels and either sex or CVD for all events.
JOINT CONTRIBUTIONS OF FASTING AND 2-HOUR GLUCOSE LEVELS
Table 2 presents relative
risks for quintiles of fasting (top rows) and 2-hour (bottom rows) glucose
levels in 3 models. For fasting glucose, adding a continuous measure of 2-hour
glucose to the age-, sex-, and race-adjusted model produced a better fitting
model (P<.001) and largely removed the association
between fasting glucose level and incident CVD event risk. Additional adjustments
for smoking, blood pressure, hypertension treatment, low-density lipoprotein
cholesterol level, body mass index, self-reported health, and clinical and
subclinical CVD had trivial effects on the fasting glucose level–CVD
association. Concerning 2-hour glucose level, adding a continuous measure
of fasting glucose to the age-, sex-, and race-adjusted model did not produce
a better fitting model (P = .40) and did not change
the association between 2-hour glucose level and CVD event risk. Adjusting
further for known CVD risk factors slightly diminished the predictive power
of 2-hour glucose level. When analyses excluded participants with diabetic-level
fasting glucose values, risk estimates changed by no more than 5% in any stratum.
|
|
|
|
Table 2. Adjusted Relative Hazards of Incident Cardiovascular Events
Among Nondiabetic and Untreated Diabetic Participants, Cardiovascular Health
Study, 1989 to 1998*
|
|
|
ANALYSES STRATIFIED BY FASTING GLUCOSE LEVEL
Table 3 presents the number
of incident CVD events, unadjusted event rates, and age-, race-, and sex (where
appropriate) adjusted hazard ratios for each 2-hour glucose category according
to fasting glucose strata. Among participants with normal fasting glucose
levels, the CVD event rate increased with each stratum of 2-hour glucose.
This observation generally held true for event types and sex and prevalent
CVD strata, although the numbers of person-years and events were limited in
the top 2-hour glucose strata for men and for those without CVD. In exploratory
analyses, a threshold model where cardiovascular risk increased 34% for 2-hour
glucose values of 154 mg/dL (8.5 mmol/L) or more fit the data best (HR, 1.34;
95% CI, 1.12-1.61). This model did not vary by event type, nor was an interaction
with sex and with prevalent CVD observed. Among participants with impaired
fasting glucose level, the event rate increased for each stratum of 2-hour
glucose level. These findings generally held true for event type, sex, and
prevalent CVD strata, although events were limited in the top 2-hour glucose
stratum for those without CVD. Exploratory analyses showed that a threshold
increase in CVD event risk of 83% for 2-hour glucose values of 161 mg/dL (8.9
mmol/L) or more fit the data best (HR, 1.83; 95% CI, 1.29-2.60). This model
did not vary by event type or by sex or prevalent CVD strata, although person-years
and events were sparse in some strata. The diabetic stratum of fasting glucose
level is not presented in Table 3
since, among this stratum, 73% of the person-years and 72% of the events were
in the 2-hour glucose stratum of 200 mg/dL (11.1 mmol/L) or more. This resulted
in too few person-years (n = 518) and events (n = 21) in the remaining 2-hour
glucose strata to calculate meaningful rates and age-, sex-, and race-adjusted
relative rates.
|
|
|
|
Table 3. Cardiovascular Event Rates per 1000 Person-Years and Age-
and Sex-Adjusted Relative Hazard Estimates According to Baseline Fasting and
2-Hour Serum Glucose Levels, Cardiovascular Health Study, 1989 to 1998*
|
|
|
Among the 3137 participants with normal fasting glucose levels, there
were 548 incident CVD events (17%), of which 5% were attributable to 2-hour
glucose levels of 140 mg/dL (7.8 mmol/L) or more and 1% were attributable
to 2-hour glucose levels of 200 mg/dL (11.1 mmol/L) or more in adjusted models
according to population attributable risk estimates. Among the 592 participants
with impaired fasting glucose, there were 141 incident CVD events (24%), of
which 24% were attributable to 2-hour glucose levels of 140 mg/dL (7.8 mmol/L)
or more and 12% were attributable to levels of 200 mg/dL (11.1 mmol/L) or
more in adjusted models.
Participants with isolated postchallenge hyperglycemia accounted for
329 (8%) of the 4014 participants at baseline. In age-, sex-, and race-adjusted
multivariate modeling, isolated postchallenge hyperglycemia was associated
with a 53% increase in CVD event risk relative to participants with 2-hour
glucose values of 139 mg/dL (7.7 mmol/L) or less (HR, 1.53; 95% CI, 1.22-1.92).
Similar results were found for CHD (HR, 1.55; 95% CI, 1.16-2.05) and CBD (HR,
1.65; 95% CI, 1.20-2.27), for women (HR, 1.67; 95% CI, 1.26-2.26) and men
(HR, 1.34; 95% CI, 0.94-1.91), and for participants with CVD (HR, 1.59; 95%
CI, 1.25-2.02). Isolated postchallenge hyperglycemia was not associated with
increased event risk for participants without CVD (HR, 0.78; 95% CI, 0.40-1.54).
COMMENT
In this study of older adults without treated diabetes at study entry,
both fasting and 2-hour glucose levels were individually associated with an
increased risk of incident CVD events, but only 2-hour glucose values were
predictive of CVD events in models that included both glycemia measures. Among
participants with normal and impaired fasting glucose levels, impaired glucose
tolerance and diabetic-level 2-hour glucose values were associated with an
increased risk of CVD events. These relationships were similar for CHD and
CBD outcomes, did not vary by sex, and were similar for participants with
prevalent clinical or subclinical CVD at study entry. Among study participants
without prevalent CVD, evidence of a glycemia-CVD association was limited.
These prospective data provide evidence that an elevated 2-hour glucose
level is a useful risk factor for identifying older adults who are at increased
risk of a CVD event, which occurred in 20% of the study participants after
8.5 years of follow-up. Besides predicting CVD events independent of fasting
glucose level, 2-hour glucose measurement contributed additional risk information
beyond fasting glucose information for participants with normal and impaired
fasting glucose levels. Among participants with normal fasting glucose levels,
above-normal 2-hour glucose values accounted for 1 of 20 incident CVD events,
and among those with impaired fasting glucose levels, above-normal 2-hour
glucose levels accounted for 1 of 4 events. If hyperglycemia is causally related
to CVD events, these statistics represent the number of CVD events that could
have been avoided in the cohort if this risk factor was removed. Epidemiologic
evidence supporting causality of hyperglycemia to CVD events, however, is
limited.12
Data from several prospective studies in nondiabetic adults have yielded
inconsistent findings on the relationship between hyperglycemia and CVD risk
in nondiabetic adults. The DECODE (Diabetes Epidemiology: Collaborative Analysis
of Diagnostic Criteria in Europe) Study Group recently published an article
on the association between fasting and 2-hour glucose levels and various mortality
outcomes and concluded that 2-hour glucose level was a better predictor of
CVD, CHD, and all-cause mortality than fasting glucose.13
Compared with participants with normal 2-hour glucose levels, impaired glucose
tolerance and diabetic-level 2-hour glucose values were associated with a
34% and 55% increase in risk, respectively. The findings are similar to findings
in this report for incident CVD. The Rancho Bernardo follow-up cohort reported
no significant associations between fasting or 2-hour glucose and either CHD
or cardiovascular mortality in men or in women when the highest quintile of
glucose measurement was compared with the lowest.14
Although the Rancho Bernardo findings were null for fasting and 2-hour glucose
measures, there was a significant threshold association (HR, 2.6) between
the highest quintiles of glycosylated hemoglobin and the lowest in women.
There was no association in men. The dissimilar findings from the CHS and
Rancho Bernardo data are not explained by the focus on fatal events, since
our findings did not change when analyses were restricted to fatal events.
Framingham Heart Study analysts found a sex difference in their data from
nondiabetic participants 45 to 84 years of age attending the 1969 to 1970
biennial examination. The authors reported a graded effect of casual glucose
level on fatal and nonfatal cardiovascular risk in women, but not in men,
after 2 years of follow-up.15 In the original
Rancho Bernardo cohort of nondiabetic participants, authors found a linear
relationship between fasting glucose level and CHD death in men and a threshold
effect (110 mg/dL [6.1 mmol/L]) in women.16
Other findings from cohorts of nondiabetic middle-aged adults, mainly men,
are heterogeneous and suggest various associations, including flat, J-shaped,
linear, and threshold relationships for various glucose measures and cardiovascular
outcomes.17-22
Risk estimates for isolated postchallenge hyperglycemia are more homogeneous
across studies. Findings from the follow-up cohort of the Rancho Bernardo
study indicated that women with isolated postchallenge hyperglycemia had a
160% increased risk of cardiovascular mortality, while men had no increased
risk when compared with nondiabetic participants.23
Recent data from the DECODE study demonstrated a 60% increased risk of death
among both sexes with isolated postchallenge hyperglycemia when compared with
nondiabetic subjects.24 In general, CHS data
tend to support the previous findings that isolated postchallenge hyperglycemia
is a cardiovascular risk factor and that this risk may be higher in women
than in men.
The primary strengths of this study include population-based sampling,
standardized glucose measurements to assess glycemia, extensive CVD risk factor
measurements, and complete follow-up for morbid and fatal events. Several
limitations merit discussion, however. First, only 1 measure each of fasting
and 2-hour glucose was available at baseline. This method is standard for
most epidemiologic studies in which populations and not individuals are studied.
According to ADA clinical practice guidelines, a diagnosis of diabetes requires
a confirmation of elevated glucose measures on a subsequent day.1
Since an elevated glucose measure was not confirmed by a second measure, we
cannot be certain that any glucose classification was correct and not the
result of random fluctuations in glycemia. Second, nearly 43% of people originally
contacted to participate in the CHS refused participation.5
Those who participated tended to be healthier than those who refused participation,
and this fact may limit the generalizability of the findings. Third, in some
analytic strata, there were too few events and person-years to produce sufficiently
narrow confidence limits to draw meaningful conclusions. Fourth, the cohort
averaged 73 years of age at entry, so data and results cannot be generalized
to younger populations. Last, although this study is primarily observational,
information from the annual clinic visits was shared with participants' physicians
and may have influenced the care received by the participants after the baseline
examination.25
CONCLUSIONS
Numerous cardiovascular risk factors have been identified during the
past 50 years of modern epidemiologic research, and the role of glycemia is
just now being defined. Any clinical or policy implications that may arise
from this study must be balanced with issues of patient burden and cost as
they relate to modifiable risk factors that affect health outcomes in older
adults.
Our findings indicate that both fasting and 2-hour postchallenge glucose
levels were associated with an increased risk of major incident coronary and
cerebrovascular events in older adults. Furthermore, 2-hour glucose level
was better able to identify those at risk than fasting glucose level alone.
Although a 2-hour measure of glycemia is burdensome in routine clinical practice,
it may serve a role in identifying older adults at increased risk of incident
cardiovascular events.
AUTHOR INFORMATION
Accepted for publication May 8, 2001.
This study was supported by contracts N01-HC-85079, N01-HC-85080, N01-HC-85081,
N01-HC-85082, N01-HC-85083, N01-HC-85084, N01-HC-85085, N01-HC-85086, N01-HL-35129,
and N01-HL-15103 from the National Heart, Lung, and Blood Institute and grant
R01-AG-09556 from the National Institute on Aging, Bethesda, Md.
We thank Melissa L. Anderson, MS, for her thoroughness and diligence
in verifying the statistical analyses in this article.
Participating Institutions and Principal
Investigators
Wake Forest University School of Medicine, Wake Forest
University, Winston-Salem, NC: Gregory L. Burke, MD. ECG Reading Center, Wake Forest University: Pentti Rautaharju, MD,
PhD. University of California, Davis: John Robbins,
MD, MHS. The Johns Hopkins University, Baltimore, Md:
Linda P. Fried, MD, MPH. MRI Reading Center, The Johns Hopkins
University: Nick Bryan, MD, PhD; Norm J. Beauchamp, MD. University of Pittsburgh, Pittsburgh, Pa: Lewis H. Kuller, MD. Echocardiography Reading Center (baseline), University of California,
Irvine: Julius M. Gardin, MD. Echocardiography Reading
Center (follow-up), Georgetown Medical Center, Washington, DC: John
Gottdiener, MD. Ultrasound Reading Center, New England Medical
Center, Boston, Mass: Daniel H. O'Leary, MD. Central
Blood Analysis Laboratory, University of Vermont, Burlington: Russell
P. Tracy, PhD. Pulmonary Reading Center, University of Arizona,
Tucson: Paul Enright, MD. Retinal Reading Center,
University of Wisconsin, Madison: Ron Klein, MD. Coordinating Center, University of Washington, Seattle: Richard A.
Kronmal, PhD. Project Officer, National Heart, Lung, and
Blood Institute, Bethesda, Md: Diane Bild, MD, MPH.
Corresponding author and reprints: Nicholas L. Smith, PhD, MPH, Cardiovascular
Health Research Unit, 1730 Minor Ave, Suite 1360, Seattle, WA 98101 (e-mail: nlsmith{at}u.washington.edu).
From the Departments of Epidemiology (Drs Smith, Heckbert, and Psaty),
Medicine (Dr Psaty), Biostatistics (Dr Kronmal), and Health Services (Dr Psaty),
University of Washington, Seattle; Division of Endocrinology, Kaiser Permanente
of Georgia, Atlanta (Dr Barzilay); Division of Epidemiology and Clinical Applications,
National Heart, Lung, and Blood Institute, Bethesda, Md (Drs Shaffer and Savage);
Department of Epidemiology, University of Pittsburgh, Pittsburgh, Pa (Dr Kuller);
and Epidemiology, Demography, and Biometry Program, National Institute on
Aging, Bethesda (Dr Resnick).
REFERENCES
| |
1. Report of the Expert Committee on the Diagnosis and Classification
of Diabetes Mellitus. Diabetes Care. 1997;20:1183-1197.
ISI
| PUBMED
2. European Diabetes Epidemiology Group. Glucose tolerance and mortality: comparison of WHO and American Diabetes
Association diagnostic criteria: the DECODE Study Group (Diabetes Epidemiology:
Collaborative Analysis of Diagnostic Criteria in Europe). Lancet. 1999;354:617-621.
FULL TEXT
|
ISI
| PUBMED
3. Barzilay JI, Spiekerman CF, Wahl PW, et al. Cardiovascular disease in older adults with glucose disorders: comparison
of American Diabetes Association criteria for diabetes mellitus with WHO criteria. Lancet. 1999;354:622-625.
FULL TEXT
|
ISI
| PUBMED
4. Fried LP, Borhani NO, Enright P, et al. The Cardiovascular Health Study: design and rationale. Ann Epidemiol. 1991;1:263-276.
PUBMED
5. Tell GS, Fried LP, Hermanson B, Manolio TA, Newman AB, Borhani NO. Recruitment of adults 65 years and older as participants in the Cardiovascular
Health Study. Ann Epidemiol. 1993;3:358-366.
PUBMED
6. Psaty BM, Lee M, Savage PJ, Rutan GH, German PS, Lyles M for the Cardiovascular Health Study Collaborative Research Group. Assessing the use of medications in the elderly: methods and initial
experience in the Cardiovascular Health Study. J Clin Epidemiol. 1992;45:683-692.
FULL TEXT
|
ISI
| PUBMED
7. Psaty BM, Kuller LH, Bild D, et al. Methods of assessing prevalent cardiovascular disease in the Cardiovascular
Health Study. Ann Epidemiol. 1995;5:270-277.
FULL TEXT
| PUBMED
8. Cushman M, Cornell ES, Howard PR, Bovill EG, Tracy RP. Laboratory methods and quality assurance in the Cardiovascular Health
Study. Clin Chem. 1995;41:264-270.
FREE FULL TEXT
9. Kuller L, Borhani N, Furberg C, et al. Prevalence of subclinical atherosclerosis and cardiovascular disease
and association with risk factors in the Cardiovascular Health Study. Am J Epidemiol. 1994;139:1164-1179.
FREE FULL TEXT
10. Ives DG, Fitzpatrick AL, Bild DE, et al. Surveillance and ascertainment of cardiovascular events: the Cardiovascular
Health Study. Ann Epidemiol. 1995;5:278-285.
FULL TEXT
| PUBMED
11. Rothman KJ, Greenland S. Modern Epidemiology. 2nd ed. Philadelphia, Pa: Lippincott-Raven; 1998:xiii, 737.
12. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared
with conventional treatment and risk of complications in patients with type
2 diabetes (UKPDS 33). Lancet. 1998;352:837-853.
FULL TEXT
|
ISI
| PUBMED
13. Kuizon D, Gordon SM, Dolmatch BL. Glucose tolerance and cardiovascular mortality: comparison of fasting
and 2-hour diagnostic criteria. Arch Intern Med. 2001;161:397-405.
FREE FULL TEXT
14. Park S, Barrett-Connor E, Wingard DL, Shan J, Edelstein S. GHb is a better predictor of cardiovascular disease than fasting or
postchallenge plasma glucose in women without diabetes: the Rancho Bernardo
Study. Diabetes Care. 1996;19:450-456.
ABSTRACT
15. Wilson PW, Cupples LA, Kannel WB. Is hyperglycemia associated with cardiovascular disease? the Framingham
Study. Am Heart J. 1991;121:586-590.
FULL TEXT
|
ISI
| PUBMED
16. Scheidt-Nave C, Barrett-Connor E, Wingard DL, Cohn BA, Edelstein SL. Sex differences in fasting glycemia as a risk factor for ischemic heart
disease death. Am J Epidemiol. 1991;133:565-576.
FREE FULL TEXT
17. Folsom AR, Szklo M, Stevens J, Liao F, Smith R, Eckfeldt JH. A prospective study of coronary heart disease in relation to fasting
insulin, glucose, and diabetes: the Atherosclerosis Risk in Communities (ARIC)
Study. Diabetes Care. 1997;20:935-942.
ABSTRACT
18. Balkau B, Shipley M, Jarrett RJ, et al. High blood glucose concentration is a risk factor for mortality in
middle-aged nondiabetic men: 20-year follow-up in the Whitehall Study, the
Paris Prospective Study, and the Helsinki Policemen Study. Diabetes Care. 1998;21:360-367.
ABSTRACT
19. Balkau B, Bertrais S, Ducimetiere P, Eschwege E. Is there a glycemic threshold for mortality risk? Diabetes Care. 1999;22:696-699.
FREE FULL TEXT
20. Burchfiel CM, Curb JD, Rodriguez BL, Abbott RD, Chiu D, Yano K. Glucose intolerance and 22-year stroke incidence: the Honolulu Heart
Program. Stroke. 1994;25:951-957.
ABSTRACT
21. Donahue RP, Abbott RD, Reed DM, Yano K. Postchallenge glucose concentration and coronary heart disease in men
of Japanese ancestry: Honolulu Heart Program. Diabetes. 1987;36:689-692.
ABSTRACT
22. Fuller JH, Shipley MJ, Rose G, Jarrett RJ, Keen H. Mortality from coronary heart disease and stroke in relation to degree
of glycaemia: the Whitehall study. Br Med J (Clin Res Ed). 1983;287:867-870.
23. Barrett-Connor E, Ferrara A. Isolated postchallenge hyperglycemia and the risk of fatal cardiovascular
disease in older women and men: the Rancho Bernardo Study. Diabetes Care. 1998;21:1236-1239.
ABSTRACT
24. The DECODE Study Group, for the European Diabetes Epidemiology Group. Consequences of the new diagnostic criteria for diabetes in older men
and women: DECODE Study (Diabetes Epidemiology: Collaborative Analysis of
Diagnostic Criteria in Europe). Diabetes Care. 1999;22:1667-1671.
FREE FULL TEXT
25. Smith NL, Heckbert SR, Bittner VA, et al. Antidiabetic treatment trends in a cohort of elderly people with diabetes:
the Cardiovascular Health Study, 1989-1997. Diabetes Care. 1999;22:736-742.
FREE FULL TEXT
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
A Step Backward--or Is it Forward?
Genuth and Kahn
Diabetes Care 2008;31:1093-1096.
FULL TEXT
Cardiometabolic Risk in Impaired Fasting Glucose and Impaired Glucose Tolerance: The Atherosclerosis Risk in Communities Study
Pankow et al.
Diabetes Care 2007;30:325-331.
|
