Effect of Dietary Modification on Incident Carotid Artery Disease in Postmenopausal Women
Results From the Women’s Health Initiative Dietary Modification Trial
Background and Purpose—Because the diagnosis and treatment of carotid artery disease may reduce the rate of stroke, the aim of this study was to determine whether a diet intervention was associated with incident carotid artery disease.
Methods—Participants were 48 835 postmenopausal women aged 50 to 79 years who were randomly assigned to either the intervention or comparison group in the Women’s Health Initiative Diet Modification Trial. Incident carotid artery disease was defined as an overnight hospitalization with either symptoms or a surgical intervention to improve flow.
Results—After a mean follow-up of 8.3 years from 1994 to 2005, there were 297 (0.61%) incident carotid artery events. In contrast to the comparison group, the risk of incident carotid disease did not differ from those assigned to the intervention group (hazard ratio, 1.08; 95% confidence interval, 0.9–1.4). In secondary analysis, there was no significant effect of the intervention on the risk for incident carotid disease during the 5 years of postintervention follow-up from 2005 to 2010 (hazard ratio, 1.24; 95% confidence interval, 0.9–1.7) and no significant effect during cumulative follow-up from 1994 to 2010 (hazard ratio, 1.13; 95% confidence interval, 0.9–1.4).
Conclusions—Among postmenopausal women, a dietary intervention aimed at reducing total fat intake and encouraging increased intake of fruit, vegetables, and grains did not significantly change the risk for incident carotid artery disease.
Carotid revascularization is frequently performed to prevent incident thromboembolic cerebrovascular accidents. Data from the North American Symptomatic Carotid Endarterectomy Trial and Asymptomatic Carotid Atherosclerosis Study indicate a significant reduction in fatal and nonfatal stroke among those undergoing surgical revascularization with significant stenosis of the carotid arteries.1,2
To clarify the effect of diet changes on several chronic diseases, the Women’s Health Initiative Dietary Modification Trial (WHI-DMT) reported the effect of achieving adherence to a diet low in total fat and higher in fruit, vegetables, and grains on the risk for incident cardiovascular disease (CVD) among postmenopausal women between the age of 50 and 79 years at baseline.3 After a mean follow-up of 8.1 years, the DMT intervention group did not have a significantly different rate of ischemic or hemorrhagic stroke (hazard ratio [HR], 1.02; 95% confidence interval [CI], 0.90–1.15).4
Because a diagnosis of carotid artery disease to include revascularization influences the likelihood of future stroke, we conducted a study to test the hypothesis of a significant effect of DMT on the rates of incident carotid artery disease in the WHI cohort. A significant decrease in these rates would suggest that a low-fat diet, which included higher intakes of fruits, vegetables, and grains, may reduce the need for carotid revascularization and, therefore, provide an impetus for enhanced efforts to recommend such a diet in public health policy.
The design of WHI-DMT has been described previously.3,4 In brief, 48 835 postmenopausal women aged 50 to 79 years were recruited from 40 sites around the United States from 1993 to 1998. During enrollment, 40% were randomly assigned to a low-fat (20% total kcal) dietary intervention group, whereas the remaining 60% were allocated to the usual diet comparison group (Figure 1).5 Exclusion criteria for DMT included type 1 diabetes mellitus, a history of cancer (except for nonmelanoma skin cancer in the previous 10 years), medical conditions predictive of a survival time <3 years or a high risk of lack of retention or intervention nonadherence.6 Women were also excluded if they (1) reported consumption of <600 kcal per day or >5000 kcal per day, (2) consumed a diet with <32% of total energy from fat, or (3) reported consuming ≥10 meals per week prepared outside of the home.6
At baseline, women could also be randomized to the postmenopausal hormone therapy (HT) trial. Details about eligibility and treatments for the HT trial have been published previously.7 After year 1, participants were invited to consider further randomization to calcium and vitamin D trial where they would be randomly assigned to supplement (500 mg of calcium carbonate and 200 IU of vitamin D3) or placebo. For the current report, a total of 8050 women (16.5%) participated in the HT trial and 25 210 (51.6%) were in the calcium and vitamin D trial.
All participants provided written informed consent. The WHI protocol and consent forms were reviewed and approved by the institutional review boards at all participating institutions.
The primary goal of DMT intervention was to reduce the risk of breast cancer by reducing total fat intake to 20% of total energy. Additional goals included increased vegetable and fruit intakes to ≥5 servings per day and increased grain intake to ≥6 servings per day.3 There were no additional diet intervention goals specific to cardiovascular risk reduction, nor was weight loss advocated.
Women in the intervention group participated in an intensive behavioral modification program consisting of 18 group sessions in the first year and quarterly maintenance sessions until the trial ended in 2005.6 Women randomly assigned to the comparison group were given a copy of Nutrition and Your Health: Dietary Guidelines for Americans and asked to maintain their usual diet.8 Dietary intake data for all participants were assessed using the WHI food frequency questionnaire, which was administered at baseline, year 1, and thereafter on a rotating sample of one third of participants every 3 years.
Demographic and personal characteristics, medication use, anthropometrics, and self-reported medical history were collected at baseline.6 Type 2 diabetes mellitus was defined as self-reported use of antidiabetic pills at any time or the use of injectable insulin. Hypertension was defined as self-reported use of an antihypertension medication. Smoking status was coded as former, current, or never. Physical activity was calculated by using a standardized classification system9 and based on self-reported physical activity data.
Semiannually, participants reported emergency room visits, overnight hospitalizations, and outpatient coronary revascularization procedures from their first follow-up to close-out in 2005. Medical records for potential reported outcomes were adjudicated by centrally trained physician adjudicators using standard criteria and blinded to randomization assignment.10
Incident carotid artery disease was defined as requiring an overnight hospitalization with either symptoms (relevant to carotid artery disease) or a surgical intervention to improve flow in the carotid arteries. The diagnosis was based on the presence of an overnight hospitalization and ≥1 of the following 3 criteria: (1) symptomatic disease with carotid artery disease listed on the hospital discharge summary; (2) symptomatic disease with abnormal findings (≥50% stenosis) on carotid angiogram, magnetic resonance angiogram, or Doppler flow study; (3) vascular or surgical procedure to improve flow to the ipsilateral brain. A diagnosis of incident stroke of any type censored the participant for future diagnoses of carotid artery disease. Individuals with a history of transient ischemic attack resulting in or occurring during overnight hospitalization were included in the analysis, but only the first case of incident carotid artery disease was included.
The analyses used time-to-event methods based on the intention-to-treat principle and included all randomized DMT participants, regardless of previous history of CVD. Follow-up time was censored at the time of a woman’s last documented follow-up contact, death, or incident stroke. HRs were estimated using Cox proportional hazard models stratified by age, previous history of carotid revascularization, hysterectomy status, HT randomization group, and calcium and vitamin D randomization group (time-dependent). Statistical significance by levels of 14 prespecified characteristics was based on tests of interaction between randomization group and subgroup. Tests of proportionality did not yield any evidence against the assumption of proportional hazards either for the intervention (P=0.30) or combined follow-up (P=0.60) periods.
After the intervention period ended in 2005, follow-up of 83.1% of surviving participants who provided written consent continued through September 30, 2010. A secondary analysis, which combined the intervention and postintervention periods, was conducted and was similar in design to the primary analysis. All statistical tests were 2-sided and nominal P≤0.05 regarded as significant.
During a mean (SD) of 8.3 (1.8) years of follow-up, there were 297 incident carotid artery disease events among the 48 835 women enrolled in the DMT. Nearly all (277; 93%) were not associated with documented cerebral infarction. Of the 297 cases, 231 (85%) included the criteria of a vascular or surgical procedure to improve flow to the ipsilateral brain, whereas 129 (47%) included symptomatic disease with abnormal findings (ie, ≥50% stenosis) on carotid angiogram, magnetic resonance angiogram, or Doppler flow study, and 102 (37%) included symptomatic disease with carotid artery disease listed on the hospital discharge summary. Eight-five percent of cases had 2 of these criteria, whereas 24% had all 3 criteria.
The baseline characteristics of participants by DMT arm assignment are presented in Table 1. There were no significant differences between participants randomized to the intervention and comparison groups for all of the characteristics except mean systolic blood pressure.
Figure 2 displays the Kaplan–Meier plot for the rates of incident carotid artery disease for the intervention and comparison groups. Overall, there was no significant effect of DMT intervention on the incidence of carotid artery disease (HR, 1.08; 95% CI, 0.85–1.36; P=0.54).
The effect of DMT intervention on incident carotid artery disease by selected subgroups is provided in Table 2. There were significant differences by history of carotid revascularization (P=0.004) and hypertension (P=0.07). Specifically, the HR for women with a history of carotid revascularization was ≈14-fold higher than that for women who did not report such a history (HR, 13.95 versus 1.02). The difference in HRs was not as pronounced for a history of hypertension (HR, 1.29) for those who reported a history of hypertension versus for those without hypertension (HR, 0.76).
We conducted secondary analyses using incident carotid artery disease data collected from 2005 to 2010. During this period, 63 (0.08%) and 79 (0.07%) incident carotid artery disease events occurred in the intervention and comparison groups, respectively, which was not significant (HR, 1.24; 95% CI, 0.89–1.73). Of note, there was no compelling evidence (P=0.49) of differential risk between the intervention (HR, 1.08) and postintervention periods (HR, 1.24). Consequently, an analysis combining these time periods was performed and revealed that DMT intervention was not associated with a significant effect on incident carotid artery disease (HR, 1.13; 95% CI, 0.93–1.36; P=0.22). As before, a history of carotid revascularization and hypertension was associated with differential risks of incident carotid artery disease for the combined time period (Table 3).
We also examined whether the effect of intervention varied by participants’ dietary habits achieved with the intervention and for the combined follow-up time. In this analysis, there were significant interactions between DMT assignment status and quartile of total energy intake (P=0.01), fruit and vegetable intake (P=0.007), and fiber intake (P=0.05; Table 4). As shown, the associations between total energy, fruit and vegetable and fiber intakes, and the DMT diet for incident carotid artery disease were strongest in the lowest quartile and somewhat U-shaped. However, when the models for fruit and vegetable as well as fiber intakes were adjusted for energy intake, these associations were no longer significant.
To determine whether the DMT intervention may have had differential effects on the biomarkers relevant to incident carotid artery disease, we conducted analyses on a subset (5.8%) of randomly selected participants (n=2815) that examined changes in these biomarkers by DMT randomization group at years 1, 3, and 6 of follow-up. In contrast to the comparison group, factor VIIc levels were 2.7% lower in the DMT intervention group at year 1 and remained at a similar difference for the duration of the trial (P=0.006). Likewise, low-density lipoprotein and high-density lipoprotein cholesterol levels were 2.4 and 1.1 mg/dL lower in the DMT intervention group at 1 year of follow-up (P<0.01 for both). However, although the difference in low-density lipoprotein-cholesterol levels persisted to the end of the trial period, that for high-density lipoprotein-cholesterol diminished annually with no appreciable difference by year 6. For triglycerides and serum glucose, the direction of effect at year 1 had reversed by year 6. There were no significant persistent differences for fibrinogen or lipoprotein-a.
Although the focus of DMT was not on a diet intervention shown to reduce cardiovascular risk, it was anticipated that adherence to a low-fat diet would have cardiovascular benefits.4 However, the results of our study showed no significant differences in the rate of incident carotid artery disease among the intervention and comparison groups. Moreover, assignment to the DMT intervention was suggestive of an unfavorable effect among those with a baseline history of coronary heart disease or hypertension and was substantially higher for those with both conditions. That is, among women who were both hypertensive and had a history of coronary heart disease, the DMT intervention resulted in more than a doubling of the risk for incident carotid disease (HR, 2.52; 95% CI, 1.26–5.04).
There are several potential reasons for this lack of effect. First, the DMT intervention was not designed to reduce the risk of CVD, nor was it a weight loss diet. Indeed, this intervention had mixed and modest effects on metabolic characteristics. Second, although women in the DMT intervention lost 2.2 kg more compared with the comparison group in the first year of the trial and maintained lower weight during the follow-up period,5 the difference in body fat was modest (<1%),11 and by year 6, the mean percentage body fat increased in both groups. Finally, the DMT intervention achieved only 70% of the targeted reduction in total fat intake needed to obtain the effect designed in the trial.12 Taken together, the overall beneficial effects of DMT intervention could be considered inadequate for changing the trajectory among those who may have already had significant carotid atherosclerosis.
Findings from this study are not entirely inconsistent with previous studies on the effect of diet on morbidity associated with atherosclerotic carotid artery disease. For instance, among men enrolled in the Framingham Heart Study, the risk of incident ischemic stroke decreased across increasing quintiles of total, saturated, and monounsaturated fats, but not polyunsaturated fat.13 Additionally, a recent comprehensive review found that adherence to Dietary Approaches to Stop Hypertension, Mediterranean, or prudent diet patterns was associated with a reduced risk of stroke, whereas a diet low in total fat was not found to have a protective effect.14 The finding of a protective effect from the Mediterranean diet may be particularly relevant to our findings because this diet does not recommend a substantial reduction in total fat intake.15 Indeed, the WHI-DMT has reported a trend toward reducing incident CVD among those who reached the lowest intakes of either saturated or trans fat.4 Of note, the effect of carotid revascularization is likely only to affect ischemic stroke rates. As such, comparisons to studies that used total (ischemic and hemorrhagic) stroke as the outcome are likely to be biased and probably not judicious.
On the contrary, the findings are somewhat disparate from many studies on the association between dietary factors and the extent of carotid atherosclerosis. In this regard, data from the Women’s Healthy Lifestyle Project showed that a lifestyle intervention that aimed to reduce dietary fat and caloric intake resulted in a significantly slower annual rate of progression of carotid intimal medial thickness compared with controls.16 Other trial and epidemiological data support favorable associations of a higher fiber intake and a lower total cholesterol intake on carotid atherosclerosis, as measured by intimal medial thickness.17 Conversely, a dietary intervention, randomized controlled trial that resulted in significant weight loss and a greater decrease in systolic blood pressure led to a regression of carotid atherosclerosis.18 Moreover, there are studies that suggest an inverse association between specific fatty acids and carotid intimal medial thickness,19 which is consistent with the hypothesis of a protective effect of certain types of dietary fat and a lower rate of stroke.20 More studies are needed to clarify the potential disparities between these findings.
An important gap in the literature is the effect of dietary intervention on incident carotid artery disease, sans incident stroke or measures of subclinical carotid atherosclerosis. Indeed, although there have been studies on the effects of pharmacological interventions on the risk for incident carotid revascularization,21 a thorough review of dietary literature did not provide any reports on this topic.
Strengths of this study include a large sample size in a well-characterized cohort that was followed for an extended period of time after a randomized intervention. In addition, the women enrolled in the WHI were all postmenopausal and relatively healthy. The latter may have resulted in a smaller number of incident carotid artery disease cases than would be expected in other populations. As such, inferences to populations dissimilar to the WHI should be made with caution. Also, in addition to those who were asymptomatic and underwent carotid artery revascularization, the outcome definition may have included some individuals with transient ischemic attack who were hospitalized overnight and that may or may not have been accompanied by a revascularization procedure. As such, this definition is somewhat heterogeneous and likely different from other studies that have examined revascularizations only.
The results of our study are similar to those obtained for the WHI-DMT intervention on incident CVD, including stroke. As such, the findings of the current analysis indicate that the effect of DMT intervention on the rates of carotid artery revascularization is not a likely explanation for the null effect of DMT intervention on incident stroke.
Sources of Funding
The Women’s Health Initiative program is funded by the National Institutes of Health through contracts HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, HHSN268201100004C, and HHSN271201100004C.
- Received February 6, 2014.
- Revision received March 14, 2014.
- Accepted March 20, 2014.
- © 2014 American Heart Association, Inc.
- 8.↵US Department of Health and Human Services. Dietary Guidelines for Americans. 4th ed. Washington, DC: USDA; 1995.
- Carty CL,
- Kooperberg C,
- Neuhouser ML,
- Tinker L,
- Howard B,
- Wactawski-Wende J,
- et al
- Sherzai A,
- Heim LT,
- Boothby C,
- Sherzai AD
- Willett WC,
- Sacks F,
- Trichopoulou A,
- Drescher G,
- Ferro-Luzzi A,
- Helsing E,
- et al
- Wu H,
- Dwyer KM,
- Fan Z,
- Shircore A,
- Fan J,
- Dwyer JH
- Shai I,
- Spence JD,
- Schwarzfuchs D,
- Henkin Y,
- Parraga G,
- Rudich A,
- et al
- Fung TT,
- Rexrode KM,
- Mantzoros CS,
- Manson JE,
- Willett WC,
- Hu FB
- Sillesen H,
- Amarenco P,
- Hennerici MG,
- Callahan A,
- Goldstein LB,
- Zivin J,
- et al