Methods

The National Heart, Lung, and Blood Institute sponsored the Women’s Angiographic and Vitamin and Estrogen (WAVE) multicenter, randomized trial, designed to test the effect of hormone replacement therapy and antioxidant vitamins on the progression of atherosclerosis measured by quantitative coronary angiography (QCA) in postmenopausal women ( ClinicalTrials.gov identifier NCT00000555 ). After 3 years of careful follow-up, the study failed to show an effect of either intervention.

Women were recruited from July 1997 to August 1999 at 7 clinical sites in the United States and Canada. To be included, women had to (1) be >55 years of age, (2) have undergone bilateral oophorectomy, or (3) have a follicle-stimulating hormone level ≥40 mlU/ml. They further had to have undergone clinically driven coronary angiography showing ≥1 coronary lesion of 15% to 75% in a major coronary artery or a primary branch within 4 months of study entry. Intervention on this lesion was not allowed. Patients were excluded if they had uncontrolled DM or hypertension, creatinine levels >2.0 mg/dl, New York Heart Association functional class IV heart failure, or a recent MI within 4 weeks of study. Supplementary Table 1 lists all exclusion criteria.

Of screened patients, 423 were enrolled. Scheduled exit coronary angiograms were obtained in 320 (76%). Intercurrent cardiovascular events and refusal of the follow-up study accounted for the others. Primary source documents describing all intercurrent events were used by an independent clinical events committee made up of 4 investigators to determine the nature of the events.

The available data set from this trial provides information on baseline and follow-up clinical status. Blood analysis was performed at baseline and at ≥1 follow-up visit. Follow-up visits were scheduled at 24, 30, 36, or 42 months. Adverse events during the study period, including cardiac events, revascularization, death, MI, and hospitalizations, were also recorded. The clinical and quantitative coronary angiographic data recorded in the original investigation formed the basis for this analysis.

QCA was performed by an independent laboratory (Stanford University, Edwin L. Alderman, MD, director). Details of the conduct of QCA and the quality assurance methods used are provided elsewhere. Measurements were carried out on up to 10 proximal epicardial artery segments. Study-driven exit coronary angiography was performed in 320 patients a mean of 2.86 years (median 3.02) from the initial angiographic study. Care was taken to obtain as nearly identical views of the segments evaluated at entry as possible. At the baseline angiographic study, segments containing no lesion of ≤15% were considered normal. If on the exit angiogram, such a segment contained a coronary narrowing of >15%, it was considered a new lesion.

For analysis, coronary segments were grouped on the basis of the patients’ baseline HbA _1c . Minimum luminal diameter (MLD) and average luminal diameter (ALD) for each segment studied were treated as continuous variables and regarded as outcome variables. Change in MLD was defined as (MLD at follow-up) − (MLD at baseline). Change in ALD was defined as (ALD at follow-up) − (ALD at baseline). These changes were compared across the strata of HbA _1c using analysis of variance. A p value for linear trend was used.

The number and percentage of new lesions (narrowing of the coronary lumen by ≥15%) appearing in segments that were normal at baseline were regarded as a second outcome variable. Differences were tested for significance using contingency tables or Fisher’s exact test. The Cochran-Armitage trend test was used to analyze the relation between HbA _1c stratum and the development of new lesions. Whether a history of DM influenced the relation between HbA _1c and the progression of coronary narrowing or the development of new lesions were evaluated using Pearson’s correlation coefficient.

Results

Potentially important baseline differences between participants who reported DM at baseline and those who did not existed. Women with known DM were slightly younger (p = 0.054) and had a higher mean weight (p <0.001). They had on average greater hip (p <0.001) and waist (p <0.001) circumferences. Moreover, women with DM more often gave histories of heart failure (p = 0.007), previous MI (p <0.05), and exercise-induced chest discomfort (p = 0.07). They more frequently were receiving angiotensin-converting enzyme inhibitors (p <0.001) and miscellaneous antihypertensive agents (p = 0.006). Total cholesterol and low-density lipoprotein levels were similar between women with DM and those without. Those with DM had lower levels of high-density lipoprotein (p <0.001) and apolipoprotein A (p <0.001) and higher levels of serum triglycerides (p = 0.009). Levels of serum creatinine (p = 0.012) and C-reactive protein (p = 0.018) levels were significantly higher in women with DM.

Table 1 displays the same characteristics compared by strata of HbA _1c level. These strata were established in line with the criteria established by the American Diabetes Association. It is thus not surprising that 93.1% of women in the highest HbA _1c strata (≥6.5%) reported histories of DM, while only 6.8% of those in the lowest strata (<5.7%) were aware of such a diagnosis. In the intermediate strata (between 5.7% and 6.5%), designated by the American Diabetes Association as “pre-diabetes,” 28.0% reported an awareness of DM. Not surprisingly, the baseline characteristics of those with HbA _1c ≥6.5% differed from those in the lowest tier (<5.7%) in a nearly identical manner to the differences between patients with known DM and those without. Moreover, it can reasonably be suggested that women who denied having received diagnoses of DM at baseline but whose baseline HbA _1c was in the pre-DM (n = 23) or DM (n = 6) range by American Diabetes Association criteria could be assumed to have recent onset of their glucose intolerance.

QCA demonstrated that women with known histories of DM (n = 154), compared to those without histories of DM (n = 267), had significantly greater decreases in MLD and ALD of coronary segments (ΔMLD −0.13 vs −0.02 mm, p <0.001; ΔALD −0.9 vs +0.01 mm, p <0.001). In addition, in women who reported DM at baseline, a higher percentage of initially nondiseased coronary segments developed new lesions during the study period (30.3% vs 17.8%, p <0.001).

The progression of coronary narrowing was then examined on the basis of baseline HbA _1c strata. These results are listed in Table 2 and shown in Figure 1 . Baseline MLD and ALD of all examined segments were similar across the 3 strata of HbA _1c , but the degree of narrowing between baseline and exit angiograms for MLD and ALD was significantly greater in segments of patients with HbA _1c ≥6.5%. Moreover, lesions were found in segments that were normal at baseline much more frequently in the higher HbA _1c strata (p <0.001; Figure 2 ). The frequency of new lesions increased incrementally from one HbA _1c stratum to another.

Graphical representation of change in QCA-measured MLD and ALD of coronary segments on the basis of subjects’ HbA _1c levels.

Graphical representation of percentage of new lesions that developed at follow-up in coronary segments on the basis of HbA _1c level. The p value for trend was <0.001.

Because nearly all patients with baseline HbA _1c ≥6.5% had clinically diagnosed DM, we assessed the possibility that the significant changes in coronary narrowing observed in the higher HbA _1c strata might simply reflect the larger number of patients with established DM in those strata. Table 3 and Figures 3 and 4 demonstrate that the significant decreases in MLD and ALD observed in the higher HbA _1c strata were driven by histories of DM and were not apparent in patients without this diagnosis. Thus, the association between HbA _1c levels and the rate of coronary narrowing appeared to be absent in participants in whom the existence of abnormal glucose metabolism was not known before WAVE study entry, presumably DM of recent onset. Importantly, even in the absence of a clinically recognized glucose abnormality there was an incremental increase in the frequency of new lesions in the higher HbA _1c strata ( Figure 5 ). It seems possible that hyperglycemia has a stronger role in the initiation of new stenosis than in the progression of established coronary narrowing.

Table 3

Minimum luminal diameter and average luminal diameter at baseline and follow-up by glycosylated hemoglobin and history of diabetes mellitus

Variable	HbA 1c <5.7%		p Value	5.7% ≤HbA 1c >6.5%		p Value	HbA 1c ≥6.5%		p Value
History of diabetes (number of lesions)	No (542)	Yes (30)		No (291)	Yes (65)		No (91)	Yes (465)
MLD at baseline (mm)	2.2	2.21	0.92	2.16	2.09	0.55	2.12	2.16	0.65
MLD at follow-up (mm)	2.17	2.1	0.62	2.16	1.92	0.03 ∗	2.12	2.02	0.22
Change in MLD (mm)	−0.04	−0.11	0.26	0	−0.18	0.002 ∗	0.01	−0.13	0.009 ∗
ALD at baseline (mm)	2.59	2.65	0.65	2.58	2.54	0.68	2.51	2.59	0.31
ALD at follow-up (mm)	2.57	2.55	0.88	2.61	2.39	0.03 ∗	2.49	2.48	0.45
Change in ALD (mm)	−0.02	−0.09	0.23	0.03	−0.15	<0.001 ∗	0.03	−0.1	0.004 ∗

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Effect of Hurricane Katrina on Chronobiology at Onset of Acute Myocardial Infarction During the Subsequent Three Years
2. Effect on Door-to-Balloon Time of Immediate Transradial Percutaneous Coronary Intervention on Culprit Lesion in ST-Elevation Myocardial Infarction Compared to Diagnostic Angiography Followed by Primary Percutaneous Coronary Intervention
3. Predictors of Mortality in Patients With Cardiovascular Implantable Electronic Device Infections
4. Prevalence, Consequences, and Implications for Clinical Trials of Unrecognized Myocardial Infarction

Stay updated, free articles. Join our Telegram channel

Join