The Journal of Nutrition Nutrition and Disease
High Fish plus Fish Oil Intake Is Associated with Slightly Reduced Risk of Venous Thromboembolism: The Tromsø Study1,2 Ida J. Hansen-Krone,3 Kristin F. Enga,3 Julie M. Su¨dduth-Klinger,4 Ellisiv B. Mathiesen,5,6 Inger Njølstad,7 Tom Wilsgaard,7 Steven Watkins,8 Sigrid K. Brækkan,3,9* and John-Bjarne Hansen3,9
Abstract Current knowledge of the effect of fish consumption on risk of venous thromboembolism (VTE) is scarce and diverging. Therefore, the purpose of the present study was to investigate the impact of fish consumption and fish oil supplements on the risk of VTE in a population-based cohort. Weekly intake of fish for dinner and intake of fish oil supplements during the previous year were registered in 23,621 persons aged 25–97 y who participated in the Tromsø Study from 1994 to 1995. Incident VTE events were registered throughout follow-up (31 December 2010). Cox-regression models were used to calculate HRs for VTE, adjusted for age, body mass index, sex, triglycerides, HDL cholesterol, physical activity, and education level. During a median of 15.8 y of follow-up there were 536 incident VTE events. High fish consumption was associated with a slightly reduced risk of VTE. Participants who ate fish $3 times/wk had 22% lower risk of VTE than those who consumed fish 1–1.9 times/wk (multivariable HR: 0.78; 95% CI: 0.60, 1.01; P = 0.06). The addition of fish oil supplements strengthened the inverse association with risk of VTE. Participants who consumed fish $3 times/wk who additionally used fish oil supplements had 48% lower risk than those who consumed fish 1–1.9 times/wk but did not use fish oil supplements (HR: 0.52; 95% CI: 0.34, 0.79; P = 0.002). In conclusion, a high weekly intake ($3 times/wk) of fish was associated with a slightly reduced risk of VTE, and the addition of fish oil supplements strengthened the inverse effect. J. Nutr. 144: 861–867, 2014.
Introduction Focus on the potential beneficial effect of dietary intake of n–3 PUFAs on cardiovascular diseases (CVDs)10 and mortality was initiated when a low rate of arterial CVD among Greenland Eskimos was related to a diet high in long-chain n–3 PUFAs of marine origin and low in saturated fat (1,2). Meta-analyses of later prospective cohort studies supported the notion that diets enriched with n–3 PUFAs protected against CVD mortality and incident stroke (3,4). Furthermore, dietary recommendations of moderate intake of fatty fish to patients suffering from acute myocardial infarction (AMI) resulted in reduction of all-cause (29%) and CVD-related (32%) mortality, but not recurrence, compared with patients not advised to consume fish (5). Similarly,
1 S. K. Brækkan and J.-B. Hansen received grants from the North Norwegian Regional Health Authority. 2 Author disclosures: I. J. Hansen-Krone, K. F. Enga, J. M. Su¨dduth-Klinger, E. B. Mathiesen, I. Njølstad, T. Wilsgaard, S. Watkins, S. K. Brækkan, and J.-B. Hansen, no conflicts of interest. 10 Abbreviations used: AMI, acute myocardial infarction; CVD, cardiovascular disease; TF, tissue factor; VTE, venous thromboembolism. * To whom correspondence should be addressed. E-mail: [email protected].
dietary supplementation with 1 g/d of highly purified n–3 PUFAs to patients that had suffered from AMI in the GISSI Study (6) resulted in a 15% risk reduction of the combined primary endpoint of death, nonfatal AMI, and stroke, but the effect was attributed to all-cause and CVD-related mortality. Experimental studies have indicated that the inverse effect of n–3 PUFA intake on CVD may be mediated by the antiarrhythmic, antithrombotic, and hypolipemic properties of n–3 PUFAs (7,8). Dietary metabolic studies have shown, however, that the antithrombotic effect of n–3 PUFAs may be counteracted by high intake of saturated fat (9). The incidence of postoperative venous thromboembolic events declined markedly during World War II (1940–1945) in Norway (10) in parallel with dietary alterations (i.e., lowered saturated fat intake and higher fish consumption) because of food rationing (11). This observation encouraged studies to explore the relation between fish consumption and risk of venous thromboembolism (VTE). An American cohort of middle-aged adults (12) reported an inverse but nonlinear relation (threshold effect) between fish consumption and risk of VTE. However, fish consumption either failed to show an association or increased the
ã 2014 American Society for Nutrition. Manuscript received December 19, 2013. Initial review completed January 30, 2014. Revision accepted March 13, 2014. First published online April 17, 2014; doi:10.3945/jn.113.189548.
861
Downloaded from jn.nutrition.org at SUNY DOWNSTATE MEDICAL CENTER on October 10, 2014
3 Hematological Research Group, Department of Clinical Medicine, University of Tromsø, Tromsø, Norway; 4Tethys Bioscience, Inc., Emeryville, CA; 5Cerebrovascular Research Group, Department of Clinical Medicine, University of Tromsø, Tromsø, Norway; 6 Department of Neurology, University Hospital of North Norway, Tromsø, Norway; 7Department of Community Medicine, University of Tromsø, Tromsø, Norway; 8Lipomics, a Division of Metabolon, West Sacramento, CA; and 9Division of Internal Medicine, University Hospital of North Norway, Tromsø, Norway
risk of VTE in cohorts of middle-aged health professionals (13) and elderly women (14), respectively. There are no obvious reasons for the apparent conflicting results from prospective cohorts on the association between fish intake and risk of VTE. However, selective populations, generally low frequencies of fish servings per week, no monitoring of additional n–3 PUFA supplementation, and lack of validation of self-reported fish intake may all contribute to the varying results. Therefore, the aim of our study was to assess the impact of fish consumption, monitored as fish servings per week with and without n–3 PUFA supplementation, on the incidence of VTE in a large prospective cohort recruited from a general population.
Methods
Assessment of fish consumption and fish oil supplements. Baseline information of the participantsÕ weekly consumption of fish for dinner and intake of fish oil supplements was assessed through a selfadministered questionnaire. For participants aged
High Fish plus Fish Oil Intake Is Associated with Slightly Reduced Risk of Venous Thromboembolism: The Tromsø Study1,2 Ida J. Hansen-Krone,3 Kristin F. Enga,3 Julie M. Su¨dduth-Klinger,4 Ellisiv B. Mathiesen,5,6 Inger Njølstad,7 Tom Wilsgaard,7 Steven Watkins,8 Sigrid K. Brækkan,3,9* and John-Bjarne Hansen3,9
Abstract Current knowledge of the effect of fish consumption on risk of venous thromboembolism (VTE) is scarce and diverging. Therefore, the purpose of the present study was to investigate the impact of fish consumption and fish oil supplements on the risk of VTE in a population-based cohort. Weekly intake of fish for dinner and intake of fish oil supplements during the previous year were registered in 23,621 persons aged 25–97 y who participated in the Tromsø Study from 1994 to 1995. Incident VTE events were registered throughout follow-up (31 December 2010). Cox-regression models were used to calculate HRs for VTE, adjusted for age, body mass index, sex, triglycerides, HDL cholesterol, physical activity, and education level. During a median of 15.8 y of follow-up there were 536 incident VTE events. High fish consumption was associated with a slightly reduced risk of VTE. Participants who ate fish $3 times/wk had 22% lower risk of VTE than those who consumed fish 1–1.9 times/wk (multivariable HR: 0.78; 95% CI: 0.60, 1.01; P = 0.06). The addition of fish oil supplements strengthened the inverse association with risk of VTE. Participants who consumed fish $3 times/wk who additionally used fish oil supplements had 48% lower risk than those who consumed fish 1–1.9 times/wk but did not use fish oil supplements (HR: 0.52; 95% CI: 0.34, 0.79; P = 0.002). In conclusion, a high weekly intake ($3 times/wk) of fish was associated with a slightly reduced risk of VTE, and the addition of fish oil supplements strengthened the inverse effect. J. Nutr. 144: 861–867, 2014.
Introduction Focus on the potential beneficial effect of dietary intake of n–3 PUFAs on cardiovascular diseases (CVDs)10 and mortality was initiated when a low rate of arterial CVD among Greenland Eskimos was related to a diet high in long-chain n–3 PUFAs of marine origin and low in saturated fat (1,2). Meta-analyses of later prospective cohort studies supported the notion that diets enriched with n–3 PUFAs protected against CVD mortality and incident stroke (3,4). Furthermore, dietary recommendations of moderate intake of fatty fish to patients suffering from acute myocardial infarction (AMI) resulted in reduction of all-cause (29%) and CVD-related (32%) mortality, but not recurrence, compared with patients not advised to consume fish (5). Similarly,
1 S. K. Brækkan and J.-B. Hansen received grants from the North Norwegian Regional Health Authority. 2 Author disclosures: I. J. Hansen-Krone, K. F. Enga, J. M. Su¨dduth-Klinger, E. B. Mathiesen, I. Njølstad, T. Wilsgaard, S. Watkins, S. K. Brækkan, and J.-B. Hansen, no conflicts of interest. 10 Abbreviations used: AMI, acute myocardial infarction; CVD, cardiovascular disease; TF, tissue factor; VTE, venous thromboembolism. * To whom correspondence should be addressed. E-mail: [email protected].
dietary supplementation with 1 g/d of highly purified n–3 PUFAs to patients that had suffered from AMI in the GISSI Study (6) resulted in a 15% risk reduction of the combined primary endpoint of death, nonfatal AMI, and stroke, but the effect was attributed to all-cause and CVD-related mortality. Experimental studies have indicated that the inverse effect of n–3 PUFA intake on CVD may be mediated by the antiarrhythmic, antithrombotic, and hypolipemic properties of n–3 PUFAs (7,8). Dietary metabolic studies have shown, however, that the antithrombotic effect of n–3 PUFAs may be counteracted by high intake of saturated fat (9). The incidence of postoperative venous thromboembolic events declined markedly during World War II (1940–1945) in Norway (10) in parallel with dietary alterations (i.e., lowered saturated fat intake and higher fish consumption) because of food rationing (11). This observation encouraged studies to explore the relation between fish consumption and risk of venous thromboembolism (VTE). An American cohort of middle-aged adults (12) reported an inverse but nonlinear relation (threshold effect) between fish consumption and risk of VTE. However, fish consumption either failed to show an association or increased the
ã 2014 American Society for Nutrition. Manuscript received December 19, 2013. Initial review completed January 30, 2014. Revision accepted March 13, 2014. First published online April 17, 2014; doi:10.3945/jn.113.189548.
861
Downloaded from jn.nutrition.org at SUNY DOWNSTATE MEDICAL CENTER on October 10, 2014
3 Hematological Research Group, Department of Clinical Medicine, University of Tromsø, Tromsø, Norway; 4Tethys Bioscience, Inc., Emeryville, CA; 5Cerebrovascular Research Group, Department of Clinical Medicine, University of Tromsø, Tromsø, Norway; 6 Department of Neurology, University Hospital of North Norway, Tromsø, Norway; 7Department of Community Medicine, University of Tromsø, Tromsø, Norway; 8Lipomics, a Division of Metabolon, West Sacramento, CA; and 9Division of Internal Medicine, University Hospital of North Norway, Tromsø, Norway
risk of VTE in cohorts of middle-aged health professionals (13) and elderly women (14), respectively. There are no obvious reasons for the apparent conflicting results from prospective cohorts on the association between fish intake and risk of VTE. However, selective populations, generally low frequencies of fish servings per week, no monitoring of additional n–3 PUFA supplementation, and lack of validation of self-reported fish intake may all contribute to the varying results. Therefore, the aim of our study was to assess the impact of fish consumption, monitored as fish servings per week with and without n–3 PUFA supplementation, on the incidence of VTE in a large prospective cohort recruited from a general population.
Methods
Assessment of fish consumption and fish oil supplements. Baseline information of the participantsÕ weekly consumption of fish for dinner and intake of fish oil supplements was assessed through a selfadministered questionnaire. For participants aged
