Invited Commentary: Diet and Risk of Venous Thromboembolism—A Hard Nut to Crack

Abstract

Recently, there has been interest in determining whether diet is associated with the risk of venous thromboembolism. The article by Varraso et al. (Am J Epidemiol. 2012;175(2):114–126) published in this issue of the Journal is an important contribution to this literature. In this commentary, the author discusses the findings of Varraso et al. within the context of the existing literature and posits epidemiologic explanations for why investigators might have failed to identify strong associations between diet and venous thromboembolism.

In this issue of the Journal, Varraso et al. (1) report the results of a comprehensive assessment of the associations between diet and incident venous thromboembolism (VTE) in a sample of 129,430 male and female health professionals. Dietary intake was represented by dietary patterns (prudent diet vs. Western diet) and intakes of various food groups and nutrients. Despite the large number of VTEs reported (n = 2,892) and the use of analytic methods that attempted to reduce dietary measurement error by taking advantage of multiple dietary assessments, overall there was little suggestion of a relation between diet and VTE. Of the 14 dietary patterns, food groups, and nutrients evaluated, only vitamin E, vitamin B₆, and fiber were associated with a lower risk of VTE in the gender-pooled analysis. Higher Western dietary pattern scores and greater consumption of red and processed meats and trans fatty acids were associated with a greater risk of VTE in men but not in women. In the present commentary, prior studies that explored the relation between diet and VTE are summarized, and potential reasons why strong associations between diet and VTE have not been consistently observed are outlined.

Etiologic data from Norway gathered during World War II (1940–1944) provided the first suggestion of a relation between diet and VTE (2, 3). In recent years, possible relations have been examined in several epidemiologic studies (1, 4–19). As shown in Table 1, some studies, but not all, showed that VTE risk was inversely associated with intakes of fruits and vegetables, alcohol, vitamin E, vitamin B₆, and fiber but positively associated with adherence to a Western dietary pattern and intakes of red and processed meat and trans fatty acids. Overall, the results of Varraso et al. (1) are consistent with those of previous investigators, who have generally failed to report strong associations between diet and VTE. Given the lack of consistent and strong associations between diet and VTE, it is important to evaluate 1) the mechanisms underlying the hypothesized associations and 2) the methodological factors that might be impairing our ability to detect relations between diet and VTE if they are present.

Table 1.

Associations Between Diet and Venous Thromboembolism in Published Articles^a

First Author, Year (Reference No.)	No. of Events	Total No. of Participants	Follow-up, years	Dietary Intervention or Assessment	Type of Association by Dietary Component or Pattern
					Inverse (Protective)	Positive (Harmful)	Null
Clinical Trials
Glynn, 2007 (4)	482	39,876	Median: 10.2	Supplement vs. placebo	Vitamin E
Ray, 2007 (5)	88	5,522	Mean: 5	Supplement vs. placebo			Folic acid, vitamin B6, and vitamin B12
den Heijer, 2007 (6)	93	701	Mean: 2.5	Supplement vs. placebo			Folic acid, vitamin B6, and vitamin B12
Cohort Studies
Steffen, 2007 (7)	196	14,962	Mean: 12.5	66-item FFQ	Vitamin B6 and fruits and vegetables	Western dietary pattern and red and processed meat	Prudent dietary pattern, whole grains, fish, omega-3 fatty acids, saturated fatty acids, and folate
Lutsey, 2009 (8)	1,950	37,393	Median: 13	127-item FFQ	Alcohol		Western dietary pattern, prudent dietary pattern, meat, dairy foods, fruit, vegetables, whole grains, refined grains, fish, coffee, alcohol, regular soda, diet soda, vitamin B6, vitamin B12, folate, vitamin E, omega-3 fatty acids, and saturated fat
Varraso, 2012 (1)	2,892	129,430	Maximum: 24	Men: 131-item FFQ Women: 61-item FFQ that increased in size with time	Vitamin E, vitamin B6, and total fiber	Western dietary pattern (men), red and processed meat (men), and trans fatty acids (men)	Prudent dietary pattern, Western dietary pattern (women), fruit, vegetables, fish, red and processed meats (women), alcohol, omega-3 fatty acids, trans fatty acids (women), and vitamins K1, B6, B12, and E
Fitzgerald, 2011 (9)	738	36,411	Mean: 14.1	133-item FFQ	Alcohol		Mediterranean-style diet, DASH-style diet
Pahor, 1996 (10)	155	7,959	Mean: 6.0	Questionnaire	Alcohol
Tsai, 2002 (11)	215	19,293	Median: 8	Questionnaire	Alcohol
Glynn, 2005 (12)	358	18,662	Median: 20.1	Questionnaire	Alcoholb
Lindqvist, 2009 (13)	312	29,518	Mean: 11	Questionnaire	Alcoholb
Hansen-Krone, 2011 (14)	460	26,662	Median: 12.5	Questionnaire	Coffeeb
Enga, 2011 (15)	462	26,755	Median: 12.5	Questionnaire
Case-Control Studies
Bhoopat, 2010 (16)	97	292	N/A	FFQ	Vegetables		Fruit, meat, fish, spicy food, and alcohol
Samama, 2000 (17)	636	1,272	N/A	Questionnaire			Alcohol
Pomp, 2008 (18)	4,423	9,658	N/A	Questionnaire	Alcohol
Yamada, 2010 (19)	100	299	N/A	Questionnaire			Alcohol

Abbreviations: DASH, Dietary Approaches to Stop Hypertension; FFQ, food frequency questionnaire; N/A, not applicable.

^aMultivariate-adjusted results are reported.

^bThere was marginal statistical significance for some categories of consumption versus none, but overall the results were not robust.

Physiologically, diet might affect VTE risk by altering levels of hemostatic and fibrinolytic factors and/or by influencing an individual’s propensity for obesity. Epidemiologic associations between VTE and specific dietary components are inconsistent, but many of the food groups and nutrients with the strongest evidence for an association with VTE—including vitamin E, the B vitamins, fruits and vegetables, red and processed meat, and alcohol—act on the hemostatic system. For example, red and processed meats are high in saturated fatty acids, which have been shown to increase platelet aggregation (20). Moderate alcohol consumption might have antithrombotic effects (21–23), although it is unclear whether these effects are due to ethanol or other compounds (24) such as resveratrol, which is present in red wine (25), or hops, which are found in beer (26). However, effects of diet on hemostatic pathways have not always translated into risk of VTE in controlled trials. B vitamin supplementation did not reduce the risk of incident VTE (5) or recurrent VTE (6), despite the apparent inverse relation with homocysteine levels (27). However, supplementation with vitamin E, which has been shown to have anticoagulant effects in humans (28, 29) and animals (30, 31), was associated with a 21% lower risk of incident VTE in the Women’s Health Study clinical trial (4).

Although numerous statistically significant associations have been observed between dietary intake and hemostatic factor levels, it must be acknowledged that the magnitudes of these associations were generally very modest. Furthermore, associations between hemostatic factors and VTE incidence are also often of modest magnitude. Given this compilation of modest physiologic associations, it is unlikely that acting through hemostatic factors, diet would have a strong impact on the absolute risk of VTE.

Diet may also influence VTE risk through its relation to obesity, which is a well-established strong risk factor for VTE (32). Obesity is associated with venous stasis and also with venous damage and varicosity (11). Further, obesity may promote thrombosis through increased activity of the coagulation cascade and decreased activity of the fibrinolytic cascade (33). Particularly when thinking about the positive associations among Western dietary patterns, red and processed meat, and VTE, it is important to be cognizant of the possibility of residual confounding by total caloric intake and obesity. Caloric intake is poorly estimated by food frequency questionnaires, and estimates are known to be differential by overweight status (34).

In addition to the potential issue of small effects between diet and hemostatic factor levels, dietary measurement error may make it difficult to detect associations between diet and VTE. In many of the publications presented in Table 1, diet was assessed only once, and follow-up was often long. Given the imprecision of dietary assessment and the possibility of an individual’s diet changing over time, dietary measurement error might have muddied the results. To reduce dietary measurement error, Varraso et al. (1) utilized data from a biennial food frequency questionnaire in a cumulative average diet analytic approach. Further, to address the issue of change in dietary intake over time, they conducted sensitivity analyses in which they used the most recent dietary intake as the predictor of VTE risk. Given their careful analytic approach, dietary measurement should be less of a concern in the article by Varraso et al. (1) than in others in which the relation between diet and VTE was explored.

Another issue to keep in mind when reviewing this (and any) literature is the possibility of publication bias. Given that many of the published studies have yielded null results, one has to question whether publication bias may also be present. It is possible that some authors have chosen not to publish papers with null results.

In summary, the evidence suggesting a relation between dietary intake and the risk of VTE is relatively weak overall. The article by Varraso et al. (1) is an important contribution to this body of literature, given the large number of cases and the analytic approach, which was intended to reduce dietary measurement error. Questions still remain regarding whether diet has an important impact on VTE risk. However, given the small magnitude of the associations between diet and hemostatic factor levels, combined with the relatively weak epidemiologic findings between diet and VTE, current data do not support a major role for diet in the development of VTE. Although exploring the relations between diet and VTE is etiologically interesting, it is also important to ask: What are the potential public health and clinical implications of this work? Given that VTE is a relatively rare condition with an incidence rate of approximately 1 per 1,000 person-years (35–37), it is unlikely that public health messages will promote adherence to a healthy diet for VTE prevention, especially as the foods hypothesized to be beneficial are also advantageous for more common conditions, such as coronary heart disease and cancer. Recurrence of VTE is common, however, with approximately 8% of individuals experiencing a recurrent event within 1 year of their initial diagnosis (35) and one-third experiencing a recurrent event in the decade after their incident event (37–39). Therefore, identifying dietary factors that are effective at preventing VTE recurrence may have clinical value.

Glossary

Abbreviation

VTE

venous thromboembolism