SEARCH

SEARCH BY CITATION

Keywords:

  • competing risks;
  • diet;
  • multivariate models;
  • obesity;
  • risk factors

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Summary.  Background: Venous thromboembolism (VTE) and cardiovascular disease (CVD) share some risk factors, including obesity, but it is unclear how dietary patterns associated with reduced risk of CVD relate to risk of VTE. Objective: To compare the relationships of adherence to a Dietary Approaches to Stop Hypertension (DASH)-style diet with the risks of CVD and VTE. Patients/Methods: We confirmed by medical record review 1094 incident cases of CVD and 675 incident VTEs during a mean follow-up of 14.6 years in 34 827 initially healthy participants in the Women’s Health Study who completed at baseline a 133-item food frequency questionnaire scored for adherence to a DASH diet. We compared estimated associations of dietary patterns with CVD and VTE from proportional hazards models in a competing risk framework. Results: Initial analyses adjusted for age, energy intake and randomized treatments showed 36–41% reduced hazards of CVD among women in the top two quintiles of DASH score relative to those in the bottom quintile (Ptrend < 0.001). In multivariate analysis, women in the top two quintiles had 12–23% reduced hazards of CVD relative to women in the bottom quintile (Ptrend = 0.04). Analyses restricted to coronary events showed more variable 10–33% reduced hazards in the top two quintiles (Ptrend = 0.09). In contrast, higher DASH scores were unrelated to risk of VTE, with a 1% reduced hazard for the top vs. bottom quintile (Ptrend = 0.95). Conclusion: An apparently strong association of adherence to the DASH diet with incidence of CVD was attenuated upon control for confounding variables. Adherence to the DASH diet was not associated with risk of VTE in women.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Dietary patterns influence cardiometabolic risk factors, including obesity, hypertension, hyperglycemia, and hyperlipidemia. In particular, randomized trials have shown that the Dietary Approaches to Stop Hypertension (DASH) diet, which is high in fruits, vegetables, and low-fat dairy products, but low in saturated and total fat, reduces blood pressure and lipid levels [1–3]. Other pathways whereby the DASH diet may reduce cardiovascular risk have been described [4], and several cohort studies have found that individuals who adhere to the DASH diet have reduced risks of both coronary heart disease (CHD) and total cardiovascular disease (CVD) [5–7]. However, the extent to which apparent benefits are mediated by weight loss is unclear [8,9].

Venous and arterial thrombosis share several pathways, as CVD and venous thromboembolism (VTE) commonly occur together, and both are strongly related to older age and obesity [10,11]. These overlaps in risk factors raise the hypothesis that lifestyle modification, including the adoption of heart-healthy diets, can lower risk of VTE [12]. Also, the reduced rate of VTE seen among women who received randomized treatment with vitamin E supplements evokes the specific question of whether dietary patterns consistent with high vitamin E intake would provide similar benefits [13]. However, some primary risk factors for CVD, including smoking, hypertension, diabetes, and elevated cholesterol, have little association or equivocal associations with VTE [14–21]. Although diet strongly influences obesity, observational evidence of associations between dietary factors and VTE has been somewhat inconsistent [22–24].

To our knowledge, no study has directly evaluated whether established dietary patterns known to be associated with CVD also predict VTE. Such associations would imply viable strategies for reducing the occurrence of VTE, and indicate substantial added benefits for adherence to such heart-healthy lifestyles. Also, formal comparisons between relative risks for these endpoints that adjust for competing risks would help to clarify the extent of shared pathways between venous and arterial events. We used prospectively collected data from the 34 827 Women’s Health Study (WHS) participants who completed a detailed food frequency questionnaire (FFQ) at baseline and were followed for a mean of 14.6 years to compare relative risks of atherosclerotic and venous thrombotic events associated with established dietary patterns.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Study participants

The WHS was a randomized trial testing whether low-dose aspirin, β-carotene and vitamin E prevented CVD and cancer in 39 876 women. The randomized findings have been described previously [25–27]. Briefly, women health professionals aged 45 years or older were enrolled into the trial from 1992 to 1994, and were initially free of CHD, cerebrovascular disease, and cancer (except non-melanoma skin cancer). Before randomization, 26 779 of the participants gave a blood sample, which was used to identify factor V Leiden and prothrombin mutations [13]. At the end of the 10-year randomized treatment period, follow-up for morbidity was complete for 97.2% of participants, and follow-up for mortality was complete for 99.4% of participants [27]. Upon conclusion of the trial in 2004, observational follow-up of 33 796 women (89% survivors) began, and continues to the present day. The response rate to the 5-year observational follow-up questionnaire was 95%. All women provided written informed consent to participate, and the study was approved by the institutional review board of Brigham and Women’s Hospital in Boston, MA.

Assessment of diet and other variables

Baseline diet was assessed with the 133-item Willett semiquantitative FFQ [28]. Participants reported routine intake of each food on a scale ranging from never or < 1 time per month to six servings or more per day. We calculated a dietary pattern score on the basis of the DASH diet [1]. The DASH score, developed by Fung et al. [6], targets eight elements: high intake of fruits, vegetables, nuts, low-fat dairy products, and whole grain, and low intake of sodium, sweetened beverages, and red and processed meats. We did not consider fruit juices as a component of fruit, as high intake has been previously shown to be associated with increased risk of CVD [29–31]. We constructed a score by first assigning women to quintiles based on their intake of a specific food. Scores for fruits, vegetables, nuts, low-fat dairy products and whole grain were assigned on the basis of the woman’s quintile ranking. For sodium, sweetened beverages, and red and processed meats, lower intakes were desired, so women in the lowest quintile received a score of 5, whereas women in the highest quintile received a score of 1. We summed scores for each of the foods to obtain an overall DASH score for each participant, which ranged from 8 to 38, with a mean of 23.8.

Participants reported body weight and height at baseline. From these, we calculated body mass index (BMI), categorized via clinical cutpoints as < 22, 22–23, 23–25, 25–29 and ≥ 30 kg m−2. They also provided information at baseline on potential confounders, including age, education, smoking status, use of hormone therapy, and postmenopausal status [10–21]. Information on hormone therapy and menopausal status was updated with follow-up questionnaires, and was included as a time-varying covariate in analyses. Physical activity was assessed at baseline, and participants were categorized into approximate quartiles corresponding to activity levels of ≤ 200, 200–600, 600–1500 and > 1500 kcal per week on the basis of the method of Mora et al. [32].

Endpoint definition

Incident CVD was the primary cardiovascular endpoint of the WHS, and included fatal and non-fatal myocardial infarction (MI), fatal and non-fatal stroke (including ischemic and hemorrhagic types), and cardiovascular death. CHD was a prospectively evaluated secondary endpoint that included non-fatal MI and death attributed to CHD. The occurrence of VTE was also prospectively assessed as a secondary endpoint [13,33]. Women completed health questionnaires every 6 months during the first year of the trial and annually each subsequent year of the trial and observational period. Individuals reporting events, including next-of-kin for decedents, were asked to authorize the release of medical records to an endpoints committee of physicians, who reviewed them in a blinded manner.

We counted only CVD and VTE events confirmed by the endpoints committee on the basis of medical record indications. Unprovoked VTE was defined as deep vein thrombosis or pulmonary embolism that occurred without the presence of cancer (diagnosed either before or up to 3 months following VTE), or immobilization, trauma, or surgery within 3 months before diagnosis.

Statistical analysis

Of the 39 876 participants, we excluded from the analysis those with missing dietary data and calorie information, those with FFQs having > 70 blank items, those with energy intakes < 600 and > 3500 kcal, those with missing information on covariates, and those with a report of prior VTE and CVD at baseline. Consistent with prior analyses of these dietary patterns and CVD, we also excluded diabetics, as this diagnosis may affect diet [4,6,7]. Hierarchical numbers excluded were: 13 women with prerandomization stroke or CHD event, 1140 women with baseline VTE, 1084 women with baseline diabetes, 530 women with missing dietary information, 790 with outlying levels of dietary intake, and 1492 women with missing covariate information. Postexclusion, 34 827 women were eligible for the analyses. Age was the time scale for proportional hazards models, and person-years of follow-up were calculated from the date of randomization to the minimum time to first incident VTE or CVD event, death, loss to follow-up, or March 2010, with a mean follow-up time of 14.6 years. We did not consider VTE occurring after CVD or CVD occurring after VTE, owing to the concern that the first event could influence diet.

DASH diet scores were categorized into quintiles, and proportional hazards regression models were stratified by event type, according to the approach detailed by Lunn and McNeil [34], to assess the association between increasing score and incident CVD or VTE. Hazard ratios and their associated 95% confidence intervals were calculated, with quintile 1 forming the reference group; tests for trend used integer scores for quintile categories. Each diet score was evaluated in a separate model, and we tested and verified the validity of the proportional hazards assumption for all primary analyses. A power calculation indicated > 90% power to detect a 25% reduction in risk associated with the top quintile of intake for each outcome.

Our model assumes different associations of each covariate with VTE and CVD. A simpler model assuming equal trends across categories of diet scores for incident CVD and VTE is nested within this model, and we tested for equal association with a likelihood ratio test. Similar analyses were performed with VTE and CHD as the two endpoints.

All analyses controlled for age, total energy intake, and randomized treatment assignments. A basic model additionally adjusted for potential confounders, including educational level, alcohol intake, smoking status, number of cigarettes smoked, time-varying postmenopausal status and hormone therapy use, and physical activity. Preliminary analyses compared the relative hazards of CVD and VTE associated with higher baseline levels of BMI. Models were further adjusted for parental history of MI before age 60 years (yes/no), but concerns that 3494 women did not provide information for this variable prevented its inclusion in the basic model. Subsequent models further adjusted for intermediates of various physiologic pathways, including BMI, history of hypertension (medication or blood pressure ≥ 140/90 mmHg [yes/no]), and history of high cholesterol (cholesterol levels ≥ 240 mg dL−1 [yes/no]).

Additional analyses considered the subscales of the DASH score, using quintiles of each component. Analyses were also stratified by provoked and unprovoked VTE to assess associations between diet score and VTE event type, with participants with incident CVD censored. Subgroup analyses evaluated the relationship between diet and incident VTE in women considered to be at high risk for VTE, including those with FV Leiden or the prothrombin gene mutation. However, no significant effect modification by these baseline factors was found, so the results are not presented.

To address the effects of each potential confounder on the relationship between CVD or VTE and the DASH diet, we added the following variables one a time to an initial model controlling for age, energy intake, randomization status, and time-varying hormone therapy and postmenopausal status: smoking status, highest level of education attained, alcohol intake, physical activity, BMI, history of high cholesterol, history of hypertension, and familial history of CHD.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Women who adhered strongly to a DASH-style diet tended to be older, be postmenopausal, be hormone therapy users, be hypercholesterolemic, be frequent exercisers, and have a higher education level (Table 1). They were also less likely to be overweight, and current smokers. Women tended to consume more calories and sodium as the DASH score increased.

Table 1.   Mean or frequency of baseline health and lifestyle characteristics across quintiles of Dietary Approaches to Stop Hypertension (DASH) dietary pattern score
 Quintiles based on DASH score
1 (n = 8606)2 (n = 7757)3 (n = 5525)4 (n = 7140)5 (n = 5799)Mean, (5th, 95th percentiles) or prevalence (%)
  1. BMI, body mass index; CHD, coronary heart disease. †All values are presented as servings per day, with the exception of sodium, which is presented as milligrams per day. ‡For the DASH diet, higher intakes are given lower scores.

Participant characteristics
 Age (years)535455565655 (46, 68)
 Whites (%)949596969695
 Current smokers (%)2314108513
 Randomization status, vitamin E (%)495050505050
 Family history of CHD (%)131313111313
 History of hypertension medication  or ≥ 140/90 mmHg (%)252524242324
 History of high cholesterol ≥ 240 mg dL−1 (%)272829303129
 Licensed practical nurse/licensed veterinary nurse education level (%)19141310814
 Kilocalories of energy  expenditure (kcal week–1)62384498111341456972 (14, 3178)
Clinical characteristics
 BMI (kg m−2)26.426.125.825.525.025.8 (19.9, 35.4)
 Baseline hormone therapy use (%)364043444641
 Postmenopausal status (%)475254596154
Dietary intake
 Energy intake (kcal)165216751719178518491728 (947, 2710)
 Alcohol intake (g day–1)4.64.64.54.54.34.5 (0, 19.6)
Components of DASH scores†
 Sodium (mg)‡179318191860191519501861 (935, 3056)
 Cola‡0.560.280.190.140.070.27 (0.00, 1.20)
 Fruits0.751.211.531.892.511.50 (0.21, 3.49)
 Vegetables2.513.253.754.355.453.74 (1.09, 7.84)
 Low-fat dairy0.520.871.131.341.691.06 (0.00, 2.93)
 Nuts0.240.320.380.460.640.39 (0.00, 1.14)
 Red meat‡1.020.800.680.570.350.71 (0.07, 1.71)
 Whole grain0.781.211.491.772.331.45 (0.13, 3.77)

Event rates and relationship with BMI

Over a mean follow-up period of 14.6 ± 2.4 years (the 5th, 25th, 50th, 75th and 95th percentiles of follow-up time were 9.4, 14.4, 15.3, 15.8 and 16.1 years, respectively), 1769 confirmed cases of incident CVD or VTE occurred, including 1094 women who had CVD first and 675 women who had VTE first. The overall incidence rates of CVD, CHD and VTE were 216, 85 and 133 events per 100 000 person-years, respectively. Higher levels of BMI were associated with increased hazards of both CVD and VTE (Table S1). For example, relative to women with BMI < 22 kg m−2, women with BMI ≥ 30 kg m−2 had a relative hazard of CVD of 1.7, and a relative hazard of VTE of 2.9, adjusted for potential confounding variables.

DASH diet score

After adjustment for age, randomization status, and energy intake, hazard ratios of incident CVD and CHD decreased linearly across quintiles of the DASH score (both Ptrend < 0.001). For CVD, significant hazard reductions of 20%, 25%, 41% and 36% were observed with quintiles 2, 3, 4, and 5, respectively, relative to quintile 1 (Table 2). However, no such pattern was detected for VTE (Ptrend = 0.95). Likelihood ratio tests revealed significantly different associations between CVD or CHD and VTE with the DASH score (CVD, P < 0.001; CHD, P = 0.001).

Table 2.   Multivariable hazard ratios (95% confidence intervals) of venous thromboembolism (VTE), cardiovascular disease (CVD) and coronary heart disease (CHD) across quintiles of Dietary Approaches to Stop Hypertension (DASH) score of the 34 827 Women’s Health Study participants
 Quintiles of dash score
12345PtrendP-value for equal association with VTE*
  1. BMI, body mass index. Unless otherwise stated, values are hazard ratios (95% confidence intervalss). *P-value associated with the null hypothesis that the trend across DASH score quintiles is the same for VTE and CVD or CHD; likelihood ratio test with one degree of freedom. †Adjusted for randomization status (aspirin, vitamin E, β-carotene), age (continuous), smoking (never, past, current), time-varying postmenopausal status (premenopausal, postmenopausal, not sure), time-varying hormone therapy use (never, past, current), alcohol intake (rarely/never, 1–3 drinks per month, 1–6 drinks per week, and ≥ 1 drink per day), energy intake (quintiles), physical activity (quartiles), cigarettes per day (continuous; approximately as 10 cigarette per day increment), and highest education level (licensed practical nurse/licensed veterinary nurse, associate’s degree, bachelor’s degree, advanced degree). ‡Models were adjusted for all covariates in the basic model, and BMI (< 22, 22.1–23.0, 23.1–24.9, 25.0–29.9, and ≥ 30), history of high cholesterol, history of hypertension and family history of CHD were added one at a time. §Family history was not included in the basic model, owing to concerns about missing data and the loss of incident events. ¶Models were adjusted for all covariates in the basic model, and BMI (< 22, 22.1–23.0, 23.1–24.9, 25.0–29.9, and ≥ 30), history of high cholesterol, history of hypertension, and family history of CHD.

Average mean score (range)17.7 (8–20)22.1 (21–23)24.5 (24–25)26.9 (26–28)30.7 (29–38)  
CVD (1094 cases)292245174194189  
 Incidence rate (per 100 000 person-years)234217215186224  
 Age, energy, and randomization status1.000.80 (0.68–0.95)0.75 (0.62–0.91)0.59 (0.49–0.70)0.64 (0.53–0.78)< 0.001< 0.001
 Basic model†1.000.94 (0.79–1.11)0.94 (0.78–1.14)0.77 (0.63–0.93)0.88 (0.72–1.07)0.040.21
 Basic model plus each set of risk factors, added one group at a time‡
  BMI1.000.94 (0.79–1.12)0.95 (0.78–1.15)0.78 (0.64–0.95)0.91 (0.74–1.10)0.080.11
  History of high cholesterol1.000.94 (0.79–1.11)0.94 (0.77–1.14)0.77 (0.63–0.93)0.87 (0.71–1.07)0.030.20
  History of hypertension1.000.93 (0.79–1.11)0.95 (0.78–1.16)0.79 (0.65–0.95)0.91 (0.74–1.11)0.090.29
  Family history of CHD§1.000.88 (0.73–1.06)0.97 (0.79–1.19)0.77 (0.62–0.94)0.83 (0.67–1.04)0.040.24
  All of the above¶1.000.88 (0.73–1.06)0.98 (0.80–1.21)0.79 (0.64–0.96)0.87 (0.70–1.09)0.110.18
CHD (430 cases)123104686471  
 Incidence rate (per 100 000 person-years)9892846184  
 Age, energy, and randomization status1.000.82 (0.63–1.06)0.71 (0.52–0.95)0.47 (0.35–0.64)0.59 (0.44–0.80)< 0.0010.001
 Basic model†1.001.00 (0.77–1.30)0.94 (0.69–1.28)0.67 (0.49–0.92)0.90 (0.65–1.24)0.090.19
 Basic model plus each set of risk factors, added one group at a time‡
  BMI1.001.00 (0.77–1.30)0.94 (0.69–1.28)0.68 (0.50–0.94)0.95 (0.69–1.31)0.160.13
  History of high cholesterol1.001.00 (0.77–1.30)0.93 (0.69–1.27)0.66 (0.48–0.91)0.89 (0.65–1.23)0.070.17
  History of hypertension1.001.00 (0.76–1.30)0.95 (0.70–1.29)0.69 (0.50–0.94)0.93 (0.68–1.29)0.140.25
  Family history of CHD§1.000.90 (0.67–1.20)0.95 (0.69–1.31)0.66 (0.48–0.93)0.83 (0.58–1.17)0.070.19
  All of the above¶1.000.89 (0.67–1.19)0.96 (0.70–1.32)0.68 (0.49–0.95)0.87 (0.61–1.22)0.130.15
VTE (675 cases)142148111152122  
 Incidence rate (per 100 000 person-years)114131137146145  
 Age, energy, and randomization status1.001.06 (0.84–1.34)1.07 (0.83–1.37)1.06 (0.84–1.34)1.00 (0.78–1.28)0.95 
 Basic model†1.001.05 (0.83–1.33)1.05 (0.82–1.35)1.05 (0.82–1.33)0.99 (0.76–1.28)0.95 
 Basic model plus each set of risk factors, added one group at a time‡
  BMI1.001.06 (0.84–1.34)1.08 (0.84–1.39)1.09 (0.86–1.39)1.08 (0.83–1.40)0.54 
  History of high cholesterol1.001.05 (0.83–1.33)1.05 (0.81–1.35)1.05 (0.82–1.33)0.99 (0.76–1.28)0.94 
  History of hypertension1.001.05 (0.83–1.33)1.05 (0.82–1.36)1.05 (0.83–1.33)1.00 (0.77–1.29)1.00 
  Family history of CHD§1.001.06 (0.83–1.35)1.02 (0.77–1.34)1.01 (0.78–1.31)1.00 (0.76–1.31)0.87 
  All of the above¶1.001.07 (0.84–1.37)1.04 (0.79–1.37)1.05 (0.81–1.36)1.08 (0.82–1.43)0.67 

A basic model adjusting for standard CVD risk factors attenuated the association between higher DASH scores and incident CVD and CHD. The trend test across quintiles of the DASH score remained significant for CVD (P = 0.04), but was no longer significant for CHD (P = 0.09), and the relative hazard reductions for CVD in quintiles 2–5 relative to quintile 1 now ranged from 6% to 22%. No association was observed for increasing DASH score and VTE in the basic model (Ptrend = 0.95; Fig. 1).

image

Figure 1.  (A) The basic model is adjusted for randomization status (aspirin, vitamin E, β-carotene), age (continuous), smoking (never, past, current), time-varying postmenopausal status (premenopausal, postmenopausal, not sure), time-varying hormone therapy use (never, past, current), alcohol intake (rarely/never, 1–3 drinks per month, 1–6 drinks per day, ≥ 1 drink per day), energy intake (quintiles), physical activity (quartiles), cigarettes per day (continuous; approximately 10 cigarettes per day), and highest education level attained (licensed practical nurse/licensed veterinary nurse, associates degree, bachelor’s degree, advanced degree). (B) Adjusted for all basic model covariates and body mass index (BMI) (< 22, 22–23, 25–25, 25–30 and ≥ 30). CVD, cardiovascular disease; DASH, Dietary Approaches to Stop Hypertension; VTE, venous thromboembolism.

Download figure to PowerPoint

We further adjusted for intermediates on various physiologic pathways and family history of CHD separately by adding each variable one at a time to the basic model. For both CVD and CHD, these inclusions resulted in slight additional attenuations of the apparent associations of categories of the DASH score with CVD and CHD.

No significant relationship was observed between higher DASH scores and reduced risk of VTE in either individually or fully adjusted models (all Ptrend > 0.50; Fig. 1). Tests of the null hypothesis that the trend across quintiles of the DASH score was equal for VTE and CVD or CHD were also insignificant (Table 2).

In analysis stratified by unprovoked (266 cases) and provoked (409 cases) event type, no relationship between categories of the DASH score and reduced risk of VTE was observed in any model (Table 3).

Table 3.   Multivariable hazard ratios of subtypes of venous thromboembolism (VTE) across quintiles of diet score
 Quintiles of dash score
12345Ptrend
  1. BMI, body mass index. Unless otherwise stated, values are hazard ratios (95% confidence intervals). *Adjusted for randomization status (aspirin, vitamin E, β-carotene), age (continuous), smoking (never, past, current), time-varying postmenopausal status (premenopausal, postmenopausal, not sure), time-varying hormone therapy use (never, past, current), alcohol intake (rarely/never, 1–3 drinks per month, 1–6 drinks per week, and ≥ 1 drink per day), energy intake (quintiles), physical activity (quartiles), cigarettes per day (continuous; approximately as 10 cigarette per day increment), and highest education level (licensed practical nurse/licensed veterinary nurse, associate’s degree, bachelor’s degree, advanced degree). †Categorical adjustment for BMI (< 22, 22.1–23.0, 23.1–24.9, 25.0–29.9, and ≥ 30).

Unprovoked VTE (266 cases)5364426047 
 Incidence rate (per 100 000 person-years)4257525856 
 Age, energy, and randomization status1.001.25 (0.87–1.80)1.10 (0.73–1.65)1.15 (0.79–1.67)1.05 (0.70–1.57)0.96
 Basic model*1.001.19 (0.82–1.72)1.01 (0.67–1.53)1.03 (0.70–1.52)0.91 (0.59–1.38)0.46
 Basic model + BMI†1.001.21 (0.83–1.74)1.04 (0.69–1.58)1.08 (0.73–1.59)1.00 (0.65–1.52)0.79
Provoked VTE (409 cases)8984699275 
 Incidence rate (per 100 000 person-years)7174858889 
 Age, energy, and randomization status1.000.95 (0.71–1.28)1.05 (0.76–1.43)1.01 (0.75–1.36)0.97 (0.71–1.33)0.98
 Basic model*1.000.96 (0.71–1.30)1.07 (0.78–1.48)1.05 (0.78–1.43)1.05 (0.75–1.46)0.61
 Basic model + BMI†1.000.98 (0.72–1.32)1.10 (0.80–1.52)1.10 (0.81–1.49)1.14 (0.82–1.59)0.32

Components of the DASH diet score were analyzed for potential relationships with VTE where foods with high vitamin E content (i.e. nuts and vegetables) were of specific interest (Table 4). We observed a 25% reduced risk of VTE among those with the highest fruit intake (≥ 2.1 servings per day) and a 23% reduced risk among those with the highest nut consumption (≥ 0.58 servings per day). No other component was significantly associated with VTE. After adjustment for BMI, we observed no significant relationship between any component and reduced risk of total VTE, or of unprovoked or provoked VTE in stratified models.

Table 4.   Multivariable hazard ratios (95% confidence intervals)* of cardiovascular disease (CVD) and venous thromboembolism (VTE) across quintiles of components of the Dietary Approaches to Stop Hypertension (DASH) diet score in 34 827 Women’s Health Study participants
 Quintiles of food intake
12345PtrendPequal association
  1. Unless otherwise stated, values are hazard ratios (95% confidence intervals). *Adjusted for randomization status (aspirin, vitamin E, β-carotene), age (continuous), smoking (never, past, current), time-varying postmenopausal status (premenopausal, postmenopausal, not sure), time-varying hormone therapy use (never, past, current), alcohol intake (rarely/never, 1–3 drinks per month, 1–6 drinks per week, and ≥ 1 drink per day), energy intake (quintiles), physical activity (quartiles), cigarettes per day (continuous; approximately as 10 cigarette per day increment), and highest education level (licensed practical nurse/licensed veterinary nurse, associate’s degree, bachelor’s degree, advanced degree). †P-value associated with the null hypothesis that the trend across DASH score quintiles is the same for VTE and CVD; likelihood ratio test with one degree of freedom. ‡Values collapsed into quartiles, as more than 20% of the population consumed 0 servings per day.

Fruit (servings per day)< 0.630.63–1.051.06–1.551.56–2.20≥ 2.21  
 CVD (1094 cases)1.000.94 (0.78–1.13)0.80 (0.65–0.97)0.84 (0.69–1.03)0.82 (0.67–1.01)0.040.76
 VTE (675 cases)1.000.98 (0.76–1.26)0.93 (0.73–1.20)0.94 (0.73–1.21)0.75 (0.57–0.99)0.04 
Vegetables (servings per day)< 1.951.95–2.822.83–3.763.77–5.19≥ 5.20  
 CVD1.000.94 (0.78–1.14)0.88 (0.73–1.07)0.97 (0.80–1.18)0.89 (0.72–1.09)0.410.22
 VTE1.000.88 (0.68–1.13)1.07 (0.83–1.39)1.08 (0.84–1.39)1.04 (0.79–1.36)0.37 
Low-fat dairy (servings per day)< 0.200.20–0.560.57–1.061.07–1.57≥ 1.58  
 CVD1.000.96 (0.80–1.15)0.89 (0.75–1.06)0.90 (0.73–1.11)0.91 (0.75–1.10)0.240.04
 VTE1.000.93 (0.72–1.21)1.15 (0.91–1.45)1.13 (0.86–1.47)1.18 (0.92–1.52)0.09 
Red and processed meat (servings per day)< 0.280.28–0.480.49–0.700.71–1.06≥ 1.07  
 CVD1.000.92 (0.77–1.12)1.03 (0.86–1.25)1.00 (0.83–1.21)0.99 (0.80–1.22)0.840.67
  VTE1.000.90 (0.71–1.15)0.99 (0.78–1.26)1.19 (0.94–1.50)0.78 (0.59–1.03)0.70 
Whole grains (servings per day)< 0.500.50–0.930.94–1.361.40–2.20≥ 2.21  
 CVD1.001.01 (0.84–1.21)1.03 (0.86–1.24)0.86 (0.71–1.05)0.96 (0.79–1.17)0.350.29
 VTE1.001.05 (0.81–1.35)1.31 (1.03–1.67)1.03 (0.79–1.33)1.12 (0.86–1.45)0.54 
Nuts (servings per day)< 0.130.13–0.200.21–0.340.35–0.57≥ 0.58  
 CVD1.000.93 (0.75–1.15)0.99 (0.83–1.17)0.94 (0.79–1.13)1.02 (0.85–1.23)0.940.15
 VTE1.000.89 (0.68–1.17)0.91 (0.73–1.13)0.94 (0.75–1.18)0.77 (0.61–0.98)0.08 
Sugar-sweetened beverages (servings per day)‡< 0.070.07–0.120.13–0.43≥ 0.44  
 CVD1.000.90 (0.76–1.07)1.06 (0.91–1.24)1.20 (1.01–1.43)0.120.12
 VTE1.000.84 (0.68–1.05)0.97 (0.79–1.18)0.94 (0.74–1.18)0.45 

To evaluate the impact of individual confounding variables, we added potential confounders one at a time to a model adjusting for age, randomization status, energy intake, and time-varying hormone therapy and postmenopausal status (Table 5). Control for cigarette smoking (including number of cigarettes smoked per day) had the strongest influence on estimates of the association of the DASH diet with the hazard of CVD, and control for BMI had the strongest influence on estimates of the association of the DASH diet with the hazard of VTE.

Table 5.   Hazard ratios (95% confidence intervals) of cardiovascular disease (CVD) and venous thromboembolism (VTE) across quintiles of Dietary Approaches to Stop Hypertension (DASH) diet score in 34 827 Women’s Health Study participants for covariates added one at a time
 Quintiles of DASH score
12345PtrendP-value for equal association with VTE*
  1. CHD, coronary heart disease. *P-value associated with the null hypothesis that the trend across DASH score quintiles is the same for VTE and CVD. †Also adjusted for amount of cigarettes smoked per day (continuous; approximately as 10 cigarette per day increment).

CVD (1094 cases)
 Age, energy, randomization status, hormone therapy use, and postmenopausal status1.000.80 (0.67–0.95)0.75 (0.62–0.91)0.59 (0.49–0.70)0.64 (0.53–0.77)< 0.001< 0.001
 Age, energy, randomization status, hormone therapy use and postmenopausal status plus each of the following added one at a time
  Smoking†1.000.90 (0.76–1.07)0.90 (0.74–1.09)0.73 (0.60–0.88)0.82 (0.68–1.00)0.0050.11
  Alcohol intake1.000.81 (0.68–0.96)0.76 (0.63–0.91)0.59 (0.49–0.71)0.65 (0.53–0.78)< 0.001< 0.001
  Education level1.000.82 (0.69–0.97)0.78 (0.64–0.94)0.61 (0.51–0.74)0.68 (0.56–0.82)< 0.0010.001
  Physical activity1.000.83 (0.70–0.98)0.79 (0.65–0.95)0.62 (0.51–0.75)0.69 (0.57–0.84)< 0.0010.002
  Body mass index1.000.80 (0.68–0.95)0.76 (0.63–0.92)0.60 (0.50–0.72)0.67 (0.55–0.80)< 0.001< 0.001
  History of hypertension1.000.80 (0.68–0.95)0.76 (0.63–0.92)0.60 (0.50–0.72)0.67 (0.55–0.81)< 0.001< 0.001
  History of high cholesterol1.000.80 (0.68–0.95)0.75 (0.62–0.91)0.58 (0.49–0.70)0.64 (0.53–0.77)< 0.001< 0.001
  Family history of CHD1.000.75 (0.62–0.90)0.78 (0.64–0.95)0.59 (0.48–0.71)0.61 (0.50–0.75)< 0.001< 0.001
 All of the above in one model1.000.88 (0.73–1.06)0.98 (0.80–1.21)0.79 (0.64–0.96)0.87 (0.70–1.09)0.110.18
VTE (675 cases)
 Age, energy, randomization status, hormone therapy use, and postmenopausal status1.001.06 (0.84–1.34)1.07 (0.83–1.37)1.07 (0.85–1.34)1.00 (0.78–1.29)0.94 
 Age, energy, randomization status, hormone therapy use and postmenopausal status plus each of the following added one at a time
  Smoking†1.001.05 (0.83–1.32)1.04 (0.81–1.34)1.04 (0.82–1.31)0.97 (0.76–1.25)0.84 
  Alcohol intake1.001.06 (0.85–1.34)1.07 (0.83–1.37)1.07 (0.84–1.34)1.00 (0.78–1.28)0.96 
  Education level1.001.06 (0.84–1.34)1.07 (0.83–1.37)1.06 (0.84–1.34)1.01 (0.79–1.30)0.90 
  Physical activity1.001.07 (0.85–1.35)1.08 (0.84–1.39)1.08 (0.85–1.36)1.02 (0.79–1.32)0.85 
  Body mass index1.001.08 (0.86–1.37)1.11 (0.86–1.42)1.12 (0.89–1.42)1.11 (0.87–1.43)0.35 
  History of hypertension1.001.07 (0.85–1.34)1.07 (0.84–1.38)1.07 (0.85–1.35)1.02 (0.79–1.30)0.86 
  History of high cholesterol1.001.06 (0.85–1.34)1.07 (0.83–1.37)1.07 (0.85–1.34)1.00 (0.78–1.29)0.94 
  Family history of CHD1.001.08 (0.84–1.37)1.04 (0.79–1.35)1.03 (0.81–1.32)1.01 (0.78–1.31)0.95 
 All of the above in one model1.001.07 (0.84–1.37)1.04 (0.79–1.37)1.05 (0.81–1.36)1.08 (0.82–1.43)0.67 

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Consistent with several prior studies, in this prospective evaluation of 34 827 initially healthy women, higher DASH diet scores were associated with moderately but statistically significantly reduced risks of overall CVD and CHD in age-adjusted and energy-adjusted models [1,9,35,36]. Control for additional confounding variables led to attenuation of these associations. In contrast, in spite of a strong association between higher BMI and risk of VTE, no associations were observed between DASH scores and VTE in any model. Overall, our study indicates potentially different relationships between diet and venous vs. arterial events.

To our knowledge, there have been only three prior studies that have explored dietary factors and their relationship with venous events. The LITE study found significant associations with VTE, using analyses based on nutrient, food and dietary pattern, in 197 cases of incident VTE [23]. Vitamin B6 and fruit and vegetable intake were inversely associated with incident VTE, whereas red and processed meat intake was positively associated with risk of VTE [23]. The investigators derived prudent (healthy) and Western (high fat intake, and low fruit and vegetable intake) dietary patterns, and found a Western-style dietary pattern to be associated with increased risk of VTE [23]. However, these results were not observed in a comparable analysis of 1950 incident VTEs using similar variables in the Iowa Women’s Health Study (IWHS) [22]. A hospital-based, case–control study of VTE in Thailand found greater consumption of vegetables and fish, and also greater consumption of spicy foods, in controls than in cases [24].

The current study differs from these prior studies in the methodological approaches used to create dietary patterns. Our analysis used a priori-developed scores of the DASH dietary pattern that may have more general application, but are consistent with both prudent and Western dietary patterns, in that intake of fruit, vegetables and fish is rewarded, and intake of saturated fat and red meat is penalized. Furthermore, diet was assessed multiple times in LITE, with a 66-item FFQ administered by an interviewer, which allowed for composite dietary analysis as well as greater precision through food model use during participant interviews [22]. Despite the self-reported dietary assessment of the IWHS and the WHS, both studies utilized FFQs with more comprehensive food lists: 127 items in the IWHS, and 133 items in the WHS [22]. However, it is likely that dietary misclassification occurred in all three cohorts.

Our study has several noteworthy limitations, in that data on risk factors and diet are self-reported, albeit by health professionals. The restriction to female health professionals limits the generalizability, although enhanced validity can arise from more accurate reporting of health behaviors and risk factors. Although we considered a substantial number of potential confounding variables, unmeasured confounding remains an important consideration. It is likely that changes in dietary intake occurred over the extended follow-up period. Nevertheless, the Spearman correlation between the DASH score calculated at baseline and then again 10 years later was 0.47, which is significant and consistent with correlations for cardiovascular biomarkers observed over a similar period [37]. Furthermore, in the creation of DASH scores, each dietary component was given equal weight, but it is possible that one part of the diet may contribute more heavily to disease risk than another. The use of such scores may also be problematic when they are applied to different populations, where equal scores could constitute intakes of components that vary considerably. Nonetheless, the observed moderate associations of each of the healthy diets with total CVD and CHD provide evidence of their validity. Although relationships of dietary scores with risk of CVD vs. VTE were no longer significantly different in analyses that controlled for lifestyle factors for DASH score models, the modest, non-significantly reduced risk of VTE associated with higher scores on the DASH diet was completely absent upon control for BMI.

The apparently different associations of adherence to a DASH diet with the risks of CVD and VTE may reflect differences in underlying disease pathways. Both blood pressure and lipid levels, which are favorably impacted on by the DASH diet, have uncertain relationships with the risk of VTE [10,11,18].

The strengths of our analysis lie in the use of competing risk survival analysis that allows for the evaluation of hazard ratios for both CVD and VTE simultaneously. Also, to our knowledge, this is the first study to evaluate whether established dietary patterns, with previously demonstrated associations with CVD, are also associated with VTE. The large number of confirmed incident venous events in the WHS remains another primary strength, and our study also presents dietary analyses stratified by provoked and unprovoked event types. Other strengths include the prospective design, extended follow-up, adjustment for energy expenditure with measures previously validated within this cohort, and the use of a nutritional assessment tool that has demonstrated its efficacy in assessing long-term dietary intake in prior research [38].

VTE occurs frequently, more often than CHD, during the 14+-year follow-up of our cohort, and is associated with high treatment costs and serious clinical sequelae. However, effective lifestyle approaches for its primary prevention have not been demonstrated. The shared relationship of obesity with the risks of both CVD and VTE suggests that dietary strategies to reduce weight are worthy of exploration as potential approaches to lower the risks of both event types. Although adherence to the DASH diet may provide additional protective effects independent of BMI for CVD, it had no significant relationship with VTE. Several alternative diets, tailored to the needs of individual patients, have been shown to lower weight and to favorably impact on CVD risk factors, and can be maintained over time [39]. It remains to be seen whether clinically meaningful weight loss, achieved and maintained through any of several such diets, is an effective approach to reduce the risks of both CVD and VTE.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

This work was supported by grant AG031061 from the National Institute on Aging.

Disclosure of Conflict of Interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

The authors state that they have no conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information
  • 1
    Appel LJ, Moore TJ, Obarzanek E, Vollmer WM, Svetkey LP, Sacks FM, Bray GA, Vogt TM, Cutler JA, Windhauser MM, Lin PH, Karanja N. A clinical trial of the effects of dietary patterns on blood pressure. DASH Collaborative Research Group. N Engl J Med 1997; 336: 111724.
  • 2
    Obarzanek E, Sacks FM, Vollmer WM, Bray GA, Miller ER 3rd, Lin PH, Karanja NM, Most-Windhauser MM, Moore TJ, Swain JF, Bales CW, Proschan MA; DASH Research Group. Effects on blood lipids of a blood pressure-lowering diet: the Dietary Approaches to Stop Hypertension (DASH) Trial. Am J Clin Nutr 2001; 74: 809.
  • 3
    Sacks FM, Svetkey LP, Vollmer WM, Appel LJ, Bray GA, Harsha D, Obarzanek E, Conlin PR, Miller ER 3rd, Simons-Morton DG, Karanja N, Lin PH, DASH-Sodium Collaborative Research Group. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. N Engl J Med 2001; 344: 310.
  • 4
    Fung TT, McCullough ML, Newby PK, Manson JE, Meigs JB, Rifai N, Willett WC, Hu FB. Diet-quality scores and plasma concentrations of markers of inflammation and endothelial dysfunction. Am J Clin Nutr 2005; 82: 16373.
  • 5
    de Souza RJ, Swain JF, Appel LJ, Sacks FM. Alternatives for macronutrient intake and chronic disease: a comparison of the OmniHeart diets with popular diets and with dietary recommendations. Am J Clin Nutr 2008; 88: 111.
  • 6
    Fung TT, Chiuve SE, McCullough ML, Rexrode KM, Logroscino G, Hu FB. Adherence to a DASH-style diet and risk of coronary heart disease and stroke in women. Arch Intern Med 2008; 168: 71320.
  • 7
    Forman JP, Stampfer MJ, Curhan GC. Diet and lifestyle risk factors associated with incident hypertension in women. JAMA 2009; 302: 40111.
  • 8
    Obarzanek E, Vollmer WM, Lin P-H, Cooper LS, Young DR, Ard JD, Stevens VJ, Simons-Morton DG, Svetkey LP, Harsha DW, Elmer PJ, Appel LJ. Effects of individual components of multiple behavior changes: the PREMIER trial. Am J Health Behav 2007; 31: 54560.
  • 9
    Folsom AR, Parker ED, Harnack LJ. Degree of concordance with DASH diet guidelines and incidence of hypertension and fatal cardiovascular disease. Am J Hypertens 2007; 20: 22532.
  • 10
    Tsai AW, Cushman M, Rosamond WD, Heckbert SR, Polak JF, Folsom AR. Cardiovascular risk factors and venous thromboembolism incidence: the longitudinal investigation of thromboembolism etiology. Arch Intern Med 2002; 162: 11829.
  • 11
    Glynn RJ, Rosner B. Comparison of risk factors for the competing risks of coronary heart disease, stroke, and venous thromboembolism. Am J Epidemiol 2005; 162: 97582.
  • 12
    Goldhaber SZ. Risk factors for venous thromboembolism. J Am Coll Cardiol 2010; 56: 17.
  • 13
    Glynn RJ, Ridker PM, Goldhaber SZ, Zee RYL, Buring JE. Effects of random allocation to vitamin E supplementation on the occurrence of venous thromboembolism: report from the Women’s Health Study. Circulation 2007; 116: 1497503.
  • 14
    Hansson PO, Eriksson H, Welin L, Svärdsudd K, Wilhelmsen L. Smoking and abdominal obesity: risk factors for venous thromboembolism among middle-aged men: ‘the study of men born in 1913’. Arch Intern Med 1999; 159: 188690.
  • 15
    Goldhaber SZ, Grodstein F, Stampfer MJ, Manson JE, Colditz GA, Speizer FE, Willett WC, Hennekens CH. A prospective study of risk factors for pulmonary embolism in women. JAMA 1997; 277: 6425.
  • 16
    Grady D, Wenger NK, Herrington D, Khan S, Furberg C, Hunninghake D, Vittinghoff E, Hulley S. Postmenopausal hormone therapy increases risk for venous thromboembolic disease. The Heart and Estrogen/progestin Replacement Study. Ann Intern Med 2000; 132: 68996.
  • 17
    Marik PE, Plante LA. Venous thromboembolic disease and pregnancy. N Engl J Med 2008; 359: 202533.
  • 18
    Everett BM, Glynn RJ, Buring JE, Ridker PM. Lipid biomarkers, hormone therapy and the risk of venous thromboembolism in women. J Thromb Haemost 2009; 7: 58896.
  • 19
    Pomp ER, Rosendaal FR, Doggen CJM. Alcohol consumption is associated with a decreased risk of venous thrombosis. Thromb Haemost 2008; 99: 5963.
  • 20
    Pahor M, Guralnik JM, Havlik RJ, Carbonin P, Salive ME, Ferrucci L, Corti MC, Hennekens CH. Alcohol consumption and risk of deep venous thrombosis and pulmonary embolism in older persons. J Am Geriatr Soc 1996; 44: 10307.
  • 21
    Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160: 341520.
  • 22
    Lutsey PL, Steffen LM, Virnig BA, Folsom AR. Diet and incident venous thromboembolism: the Iowa Women’s Health Study. Am Heart J 2009; 157: 10817.
  • 23
    Steffen LM, Folsom AR, Cushman M, Jacobs DR Jr, Rosamond WD. Greater fish, fruit, and vegetable intakes are related to lower incidence of venous thromboembolism: the Longitudinal Investigation of Thromboembolism Etiology. Circulation 2007; 115: 18895.
  • 24
    Bhoopat L, Rojnuckarin P, Hiransuthikul N, Intragumtornchai T. Low vegetable intake is strongly associated with venous thromboembolism in Thai population. Blood Coagul Fibrinolysis 2010; 21: 75863.
  • 25
    Cook NR, Lee I-M, Gaziano JM, Gordon D, Ridker PM, Manson JE, Hennekens CH, Buring JE. Low-dose aspirin in the primary prevention of cancer: the Women’s Health Study: a randomized controlled trial. JAMA 2005; 294: 4755.
  • 26
    Lee I-M, Cook NR, Gaziano JM, Gordon D, Ridker PM, Manson JE, Hennekens CH, Buring JE. Vitamin E in the primary prevention of cardiovascular disease and cancer: the Women’s Health Study: a randomized controlled trial. JAMA 2005; 294: 5665.
  • 27
    Ridker PM, Cook NR, Lee I-M, Gordon D, Gaziano JM, Manson JE, Hennekens CH, Buring JE. A randomized trial of low-dose aspirin in the primary prevention of cardiovascular disease in women. N Engl J Med 2005; 352: 1293304.
  • 28
    Willett WC. Nutritional Epidemiology. New York, NY: Oxford University Press, 1998.
  • 29
    Hansen L, Dragsted LO, Olsen A, Christensen J, Tjønneland A, Schmidt EB, Overvad K. Fruit and vegetable intake and risk of acute coronary syndrome. Br J Nutr 2010; 104: 24855.
  • 30
    Hung H-C, Joshipura KJ, Jiang R, Hu FB, Hunter D, Smith-Warner SA, Colditz GA, Rosner B, Spiegelman D, Willett WC. Fruit and vegetable intake and risk of major chronic disease. J Natl Cancer Inst 2004; 96: 157784.
  • 31
    Mozaffarian D, Hao T, Rimm EB, Willett WC, Hu FB. Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med 2011; 364: 2392404.
  • 32
    Mora S, Cook N, Buring JE, Ridker PM, Lee I-M. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation 2007; 116: 211018.
  • 33
    Glynn RJ, Ridker PM, Goldhaber SZ, Buring JE. Effect of low-dose aspirin on the occurrence of venous thromboembolism: a randomized trial. Ann Intern Med 2007; 147: 52533.
  • 34
    Lunn M, McNeil D. Applying Cox regression to competing risks. Biometrics 1995; 51: 52432.
  • 35
    Blumenthal JA, Babyak MA, Hinderliter A, Watkins LL, Craighead L, Lin PH, Caccia C, Johnson J, Waugh R, Sherwood A. Effects of the DASH diet alone and in combination with exercise and weight loss on blood pressure and cardiovascular biomarkers in men and women with high blood pressure: the ENCORE study. Arch Intern Med 2010; 170: 12635.
  • 36
    Panagiotakos DB, Pitsavos C, Stefanadis C. Alpha-priori and alpha-posterior dietary pattern analyses have similar estimating and discriminating ability in predicting 5-Y incidence of cardiovascular disease: methodological issues in nutrition assessment. J Food Sci 2009; 74: H21824.
  • 37
    Glynn RJ, MacFadyen JG, Ridker PM. Tracking of high-sensitivity C-reactive protein after an initially elevated concentration: the JUPITER Study. Clin Chem 2009; 55: 30512.
  • 38
    Salvini S, Hunter DJ, Sampson L, Stampfer MJ, Colditz GA, Rosner B, Willett WC. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol 1989; 18: 85867.
  • 39
    Sacks FM, Bray GA, Carey VJ, Smith SR, Ryan DH, Anton SD, McManus K, Champagne CM, Bishop LM, Laranjo N, Leboff MS, Rood JC, de Jonge L, Greenway FL, Loria CM, Obarzanek E, Williamson DA. Comparison of weight-loss diets with different compositions of fat, protein, and carbohydrates. N Engl J Med 2009; 360: 85973.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  10. Supporting Information

Table S1. Multivariable hazard ratios (95% confidence intervals) of venous thromboembolism (VTE) and cardiovascular disease (CVD) across clinical categories of body mass index (BMI) in 34 827 Women's Health Study participants.

FilenameFormatSizeDescription
JTH_4588_sm_TableS1.docx15KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.