Journal of Clinical Pharmacy and Therapeutics, 2014, 39, 56–60

doi: 10.1111/jcpt.12104

A practical approach to minimize the interaction of dietary vitamin K with warfarin C.-H. Chang* PhD, Y.-W. Wang*† MSc, P.-Y. Yeh Liu*‡ MSc and Y.-H. Kao Yang* BSPharm *Institute of Clinical Pharmacy and Pharmaceutical Sciences, College of Medicine, National Cheng Kung University, Tainan, †Department of Pharmacy, Buddhist Tzu Chi Medical Foundation, Taipei, and ‡Department of Pharmacy, National Cheng Kung University Hospital, Tainan, Taiwan

Received 13 October 2012, Accepted 23 September 2013

Keywords: diet, interaction, vitamin K, warfarin

the coagulation cascade. Not surprisingly, additional vitamin K can counteract the anticoagulant effect of warfarin.5 Therefore, irregular intake of vitamin K from vegetables will influence the therapeutic effect of warfarin.6 Vitamin K is a fat-soluble vitamin. There are three forms of vitamin K: vitamin K1 (phylloquinone), vitamin K2 (menaquinones) and vitamin K3 (menadione). Of these, vitamin K1 is the major dietary source of vitamin K, being present in relatively large amounts in dark green vegetables and certain plant oils.2,7 In the literature, there are many clinical studies examining the effects of vitamin K intake on coagulation parameters [e.g. international normalized ratio (INR)] in patients on anticoagulation medication.3,4,6,8,9 Karlson et al. found that continuous administration of vitamin K1 100 lg/day for 1 week did not significantly affect thrombotest (TT) values, but continuous administration of vitamin K1 250 lg/day for 1 week resulted in TT values outside the therapeutic range.9 Schurgers et al.8 also reported a similar result: additional vitamin K1 of 100 lg/day did not interfere with oral anticoagulant therapy, whereas intake of vitamin K 150 lg/ day led to significant change in INR. Furthermore, Khan et al.3 showed that for each 100 lg increase in vitamin K dietary intake over 4 days, the INR was reduced by 0
2. Given that the established impact of dietary vitamin K intake on oral anticoagulation, de Assis et al.10 suggested that adjusting the number of times each vitamin K-rich food item is ingested per week would help patients maintain their INR target. For example, if the vitamin K intake of a patient on anticoagulants was based on two vitamin K-rich food items (e.g. broccoli and cabbage) four times per week each, dietary modification may involve decreasing the ingestion of the same food items to two times per week. Their results showed that patients who adjusted their dietary vitamin K intake reached their target INR more frequently than those who changed their anticoagulant prescription. This study suggests that using an approximate measure of weekly vitamin K intake, instead of a daily measure, could better help patients maintain a steady target INR. Additionally, it might be more practical for patients. However, most vitamin K contents of vegetables in the literature are provided as weights of salad, which is rarely seen in daily meals in Taiwan or, indeed, most Asian countries, where vegetables are usually cooked. Unfortunately, there is no widely acceptable, patient-friendly strategy for managing this troublesome interaction, consisting of ‘maintaining constant weekly dietary vitamin K intake scores’. Therefore, there exists a need to establish a practical guideline of dietary vitamin K intake, to counsel patients receiving warfarin in Taiwan. Based on the previous research, we aimed to establish a measure of the vitamin K equivalent for vegetables common in Taiwan,

SUMMARY What is known and objective: The many interactions between warfarin and other drugs and foods generate great challenges for clinicians and patients in maintaining stable anitcoagulation. Interactions due to variable vitamin K content of different dietary items influence the therapy of nearly all patients on warfarin. Unfortunately, there is no widely acceptable, patientfriendly strategy for managing such interactions. In this contribution, we propose a practical approach to managing this troublesome interaction, consisting of ‘maintaining constant weekly dietary vitamin K intake scores’. Methods: Twenty-three vitamin K-rich vegetables commonly seen in Taiwanese meals were identified and classified into seven score grades according to their relative vitamin K content per serving. The scores were based on published vitamin K content of different foods. Results and discussion: The vitamin K score was equivalent to 5 points for spinach and garland chrysanthemum per bowel, followed by (baby) bok choy, amaranth, arden lettuce (4 points); leaf mustard, edible rape, sweet potato leaf, bai cai and Chinese leek (3 points); and okra and Chinese celery (0
5 points). This classification can be used to guide patients in recording their weekly vitamin K scores with a view to maintaining it when on warfarin. What is new and conclusion: We suggest a novel approach to patient counselling on warfarin to maintain consistent dietary vitamin K intake and achieve a more stable anticoagulation response. A prospective randomized controlled trial to validate this pragmatic approach would be useful. WHAT IS KNOWN AND OBJECTIVE For six decades, since its launch in 1950s, warfarin has remained the most widely used oral anticoagulant, despite the significant challenge which its use presents as a result of its complicated drug and food interactions.1,2 Vitamin K-rich vegetables are considered a major dietary factor contributing to the unstable effect of warfarin.3,4 Warfarin acts as an anticoagulant by inhibiting vitamin K epoxide reductase, which is an essential enzyme in Correspondence: Y.-H. Kao Yang, Institute of Clinical Pharmacy and Pharmaceutical Sciences, College of Medicine, National Cheng Kung University, No. 1 University Road, Tainan 70101, Taiwan. Tel.: +886 6 2353535 5688; fax: +886 6 2373149; e-mail: [email protected] Yu-Wen Wang and Chia-Hsien Chang contributed equally.

© 2013 John Wiley & Sons Ltd

56

C.-H. Chang et al.

Interaction of dietary vitamin K

which would provide a practical approach to manage the interaction between warfarin and vitamin K-rich vegetables. This approach may be suitable for other Asian populations. With variable extents of adjustment, and a list of vitamin K dietary equivalents, the approach may also be appropriate for an even wider range of populations.

Table 1. Vitamin K content of vegetables common in Taiwanese household meals

Vegetable items (g)

Vitamin K content per 100 ga (lg/100 g)

Weight per bowlb

Vitamin K content per bowl (lg/bowl)

Amaranth (baby) Bok choy Sweet potato leaf Spinachc Arden lettuce Leaf mustard (Mustard greens) Edible rape (Komatsuna)c Basil Broccolic Chinese leekc Bai cai (Pai tsai)c Water convolvulusc Chayote shoot Garland chrysanthemum Gynura bicolor Scallion Chinese cabbagec Cabbagec Chinese leek flower Chinese kale Asparagus Okra Chinese celery

48111 46311 46311 49814 41411 39411 31913 30211 30713 29014 27014 25014 24011 23013 22511 20712 19014 12713 12711 10811 8012 4812 4711

104 130 85 150 120 110 130 76 90 120 140 90 72 300 89 67 170 120 106 120 145 120 108

500 602 394 747 497 433 415 230 256 348 378 225 173 690 200 139 323 163 135 130 116 58 51

METHODS Vitamin K content of vegetables in Taiwan’s daily diet We searched the National Nutrient Database of Taiwan (established by the Department of Health, Executive Yuan, Taiwan), PubMed, Airiti Library (Chinese Electronic Periodical Services, CEPS) and the National Digital Library of Theses and Dissertations in Taiwan to obtain vitamin K content of common food items in Taiwan. Only Huang’s Master’s thesis11 provided the required information. In addition, the USDA (United States Department of Agriculture) National Nutrient Database for Standard Reference,12 the article published by Kamao et al.13 and the ingredient of Japanese food report from the Ministry of Education, Culture, Sports, Science and Technology Japan14 listed certain vegetables commonly seen in meals in Taiwan. These vegetables, together with their vitamin K content on a weight/weight basis are shown in Table 1. Measuring vitamin K content by bowl As rice bowls are commonly used for daily meals in Taiwan, we converted the vitamin K content from lg/100 g into lg/bowl. In this experiment, the volume of one typical household rice bowl was measured as 270 mL when filled up to the horizontal line (Figs 1 and 2). All vegetables were purchased from traditional markets in Taipei and then cooked by boiling with water. The vitamin K content of each vegetable was determined as the content of one bowl of boiled vegetable, except scallion. When vitamin K contents of vegetables in the literature were measured with cooked samples, such as okra and asparagus, we weighed one bowl of those vegetables in boiled form and then converted the vitamin K content from lg/100 g into lg/bowl. On the other hand, when the known vitamin K content was available for a raw vegetable, we weighed the raw vegetable required for making one bowl of boiled vegetable by the following steps:

a

Vitamin K contents listed were based on the weight of raw vegetables, except those of asparagus and okra, which are listed as cooked weight. For asparagus and okra, the ‘weight per bowl’ indicated the weight of one bowl of boiled vegetable; for scallion, it indicated the weight of one bowl of raw scallion; for the remaining vegetables, it indicated the weight of raw vegetable required to produce one bowl of boiled vegetable. c The data reported in Huang et al.11 are listed as following: spinach 416 lg/ 100 g, edible rape 306 lg/100 g, broccoli 284 lg/100 g, Chinese leek 271 lg/100 g, bai cai 258 lg/100 g, water convolvulus 243 lg/100 g and cabbage 144 lg/100 g. b

Diameter: 11·5 cm

1 Remove non-edible parts and wipe dry the surface of the vegetable to be tested; 2 Weigh the remaining part prior to cooking (W1); 3 Boil with water by the household cooking method and weigh the cooked vegetable after draining (W2); 4 Fill a standard household rice bowl with boiled vegetable; 5 Measure the weight of boiled vegetable in the bowl (W3); 6 The weight of raw vegetable required for making one bowl of boiled vegetable (W4) could be estimated as: W4 = W1 9 W3/W2 7 Recook the vegetable based on W4 to check whether W4 was accurate; if not, adjust and retest until the standard is matched as Fig. 2. 8 Once W4 is confirmed, the vitamin K content of one bowl of boiled vegetable is calculated as: (vitamin K content per 100 g)a 9 W4 (g)/ 100

Height: 4·8 cm

Height: 5·7 cm

Fig. 1. The bowl used in this experiment. The volume was 270 mL as filled up to the horizontal line and was identical with bowls most commonly in use in Taiwan. Top diameter: 11
5 cm; height: 5
7 cm with base and 4
8 cm without base. Calculating the relative ratio of the study vegetables’ vitamin K content

Because scallion is mainly used as condiment without cooking, its vitamin K content is presented as one bowl of raw scallion.

Using the vitamin K content of one bowl of each vegetable, we ranked the vegetables and calculated their relative ratios. Onebowl vitamin K (OVK) equivalent was defined by the median

a

Second column of Table 1.

© 2013 John Wiley & Sons Ltd

Journal of Clinical Pharmacy and Therapeutics, 2014, 39, 56–60 57

C.-H. Chang et al.

Interaction of dietary vitamin K

calculate the total intake of vitamin K in a fixed time period (e.g. 1 week or a shorter period) and helped to obtain a constant level of dietary vitamin K. For instance, when a patient reaches a stable dietary pattern, they can calculate their total vitamin K score for 1 week and then keep the weekly vitamin K score constant by recording their self-scored intake of vitamin K-rich vegetables. Stable therapeutic anticoagulation with warfarin is dependent on a balance being achieved between inhibition of the recycling enzymes and the amount of dietary vitamin K entering the cycle to support less efficient carboxylation. In this article, we propose a novel approach to deal with the interaction of vitamin K-rich foods and warfarin. Advantages of this approach include (i) explicit guidance on managing vegetable ingestion to minimize fluctuation in vitamin K intake and (ii) a practical scoring method that can be tailored to individual needs. The algorithm can be used by patients with any level of vegetable intake. In current practice, healthcare professionals advise patients on warfarin to maintain their vitamin K intake at a constant level. However, there is no quantitative reference to guide the patients. Some physicians and pharmacists counsel their patients to limit or avoid foods high in vitamin K, whereas other experts argue against this on the grounds that this may result in deficiencies of other essential vitamins and minerals contained in the same foods.3,7 By contrast, some evidence suggests that patients with a low intake of dietary vitamin K could be sensitive to even small changes in vitamin K intake, potentially impairing their anticoagulant stability.15,16 Therefore, patients on anticoagulants should keep their intake of vitamin K adequate and constant. The ‘adequate intake’ of vitamin K recommended by the US National Academy of Science is 90 lg/day for females (≥19 year) and 120 lg/day for males (≥19 year).17 It has been reported that multiple administrations of vitamin K 100 lg/day from pure vitamin preparations do not significantly interfere with oral anticoagulant therapy.8,9 The bioavailability of vitamin K from various vegetables (whether eaten raw or cooked, and with or without fat) is about 4–34% of that of pure vitamin K supplements.8,18,19 Karlson et al.9 found that a single dose of 250 lg vitamin K1 as well as an intake of 250 g broccoli or spinach did not significantly affect the warfarin effect. When advising on vegetables’ vitamin K content per bowl, the data shown in Table 1 can be used.

Fig. 2. One bowl of cooked vegetable for weighing. The standard procedure was to fill the bowl level up to the rim with cooked vegetable. listed in this study. Predefined relative ratios were set, and the vegetables were considered as equivalent within the tier if the content of vitamin K per bowl was in the range 80–120% or 95– 115% to avoid overlap. The equivalent amount of the vegetables is the reciprocal for predefined cut-off points. As a practical approach, the vitamin K score of a given vegetable was assigned double the value of the relevant relative ratio. All data were analysed with Microsoft Office Excel 2010 (Microsoft, Redmond, WA, USA). RESULTS AND DISCUSSION The vitamin K contents of 23 vegetables commonly consumed in Taiwan are available and ranged from 47 to 498 lg/100 g.11–14 The vitamin K contents for one bowl of vegetable are listed in Table 1. Spinach and garland chrysanthemum have the highest content of vitamin K per bowl, whereas okra and Chinese celery have the lowest. One vitamin K bowl equivalent is 276 lg per bowl, which is equal to one bowl of broccoli. The vitamin K equivalent of the study vegetables is presented in Table 2, using the bowl as the unit of comparison. For example, the vitamin K content of one bowl of boiled broccoli is comparable to a half bowl of boiled amaranth or two bowls of boiled asparagus. We established an algorithm to rank the vegetable’s vitamin K content by score, as shown in Table 2. This algorithm was used to

Table 2. Vitamin K equivalent of study vegetables and their corresponding vitamin K score

Relative Ratioa

Lower limit of relative ratio

Upper limit of relative ratio

Vitamin K content (lg/bowl)

Vegetable items

Equivalent amountb

Vitamin K Scorec

0
25 0
50 0
75 1
00 1
50 2
00 2
50

0
2 0
4 0
6 0
9 1
3 1
8 2
3

0
3 0
6 0
9 1
2 1
7 2
2 2
8

59–79 117–159 176–238 235–317 352–476 497–607 621–759

Okra, Chinese celery Scallion, Chinese leek flower, Chinese kale, asparagus Basil, water convolvulus, gynura bicolor, chayote shoot, cabbage Broccoli, Chinese cabbage Leaf mustard, edible rape, sweet potato leaf, bai cai, Chinese leek (baby) Bok choy, amaranth, arden lettuce Spinach, garland chrysanthemum

4 bowels 2 bowels 1 and 1/3 bowels 1 bowel 2/3 bowel 1/2 bowel 1/2 bowel

0
5 1 1
5 2 3 4 5

The content of vitamin K for all vegetables was divided by the one-bowl vitamin K equivalent of 276 lg. All vegetables in bowls were cooked except scallion. c To improve the applicability of the table, score calculation could be adapted to patients’ own eating habits, such as a half bowl or one dish, rather than relying on fixed vitamin K content per bowl. a

b

© 2013 John Wiley & Sons Ltd

Journal of Clinical Pharmacy and Therapeutics, 2014, 39, 56–60 58

C.-H. Chang et al.

Interaction of dietary vitamin K

cabbage and raw basil are 190 and 302 lg/100 g, respectively. However, to fill up a bowl with cooked vegetables, the amount of raw Chinese cabbage and raw basil by weight was 170 and 76 g, respectively. This means that less vitamin K is contained in one bowl of basil than in one bowl of Chinese cabbage, although 100 g of raw basil contain more vitamin K than 100 g of raw Chinese cabbage. In this study, we found that one bowl could be filled with 15 spears of small okra, each spear weighing around 8 g and containing 4 lg of vitamin K (Table 1). Kamao et al.13 reported that the vitamin K content of raw garland chrysanthemum was 230 lg/100 g. However, in the cooked state, this increases remarkably to 627 lg/100 g. Interestingly, garland chrysanthemum is a vegetable popularly known to displease cooks, because it shrinks heavily during cooking. When estimating vitamin K intake, how a food is consumed in different countries needs to be taken account of. There are some limitations to this study. First, estimation by filling vegetables up to the rim of a bowl is semi-quantitative relative to accurate measurements of liquids. Additionally, different growing conditions, species, ripeness and portions of leaf and stem may lead to variation in vitamin K content of the same vegetable.22 Moreover, the volume of vegetables may vary with the time allowed for draining after boiling in water. All the above factors may result in variations in vitamin K content per bowl of vegetable. However, these variations would not affect the relative ratio of vitamin K content between the vegetables and may not be of practical importance.

Several approaches have been proposed for minimizing fluctuations in dietary vitamin K intake. Some have suggested that patients to eat 1–2 servings of vitamin K-containing foods on 2– 3 days per week.7 The US National Institutes of Health recommend limiting intake of foods ‘high’ and ‘moderately high’ in vitamin K to no more than one and three servings each day, respectively.20 Compared with our algorithm, these suggestions cannot be individualized. Furthermore, they are only applicable in Western countries as only Western foods are listed. For example, the vitamin K content of (baby) bok choy, amaranth, arden lettuce, sweet potato leaf, gynura biocolor, chayote shoot, are listed only in Taiwanese publications. Moreover the content of Vitamin K for leaf mustard, basil, Chinese leek flower, Chinese kale, asparagus, okra and Chinese celery show marked geographical variability.11–14 Therefore, the vitamin K of various vegetables should be estimated based on local produce to take account of differences due to different variants grown in different geographical area and variable environmental influences. Data available in both Taiwan and Japan, suggest that differences in the vitamin K content of several vegetables were less than 15%. Thus, our report may be valid for patient counselling in Japan. Table 2 may serve as a convenient source of information for patients to stabilize their vitamin K intake. As this scoring table is based on relative vitamin K content of the different vegetables, patients can adjust their intake based on the size of their food servings (e.g. half a bowl or one dish). If a patient usually eats a fixed amount of vegetable at a time, they can employ ‘one meal’ as one serving. As vitamin K loss from cooking is negligible, we were able to estimate the vitamin K content of boiled vegetables per bowl, based on the vitamin K content of the raw vegetables per 100 g.21 Although the USDA National Nutrient Database has listed vitamin K content of cooked foods per common measure (e.g. one cup), this information needs to be adjusted for different countries to include foods commonly consumed as we have done in this study. The ranking of vegetables varied (Table 1) when the measurement basis of vitamin K content was converted from lg/100 g into lg/bowl. For example, the vitamin K content of raw Chinese

WHAT IS NEW AND CONCLUSION We provide a method for scoring the vitamin K content of vegetables rich in this vitamin. It is hoped that this tool will support health care professionals, when counselling patients on warfarin so as to maintain stable anticoagulation. A prospective randomized controlled trial to validate this pragmatic approach would be useful.

REFERENCES 1. Zehnder JL. Drugs used in disorders of coagulation. In: Katzung BG, Masters SB, Trevor AJ, eds. Drugs used in disorders of coagulation, 11th edn. New York: McGrawHill, 587–603, 2009. 2. Nutescu E, Chuatrisorn I, Hellenbart E. Drug and dietary interactions of warfarin and novel oral anticoagulants: an update. J Thromb Thrombolysis, 2011;31:326–343. 3. Khan T, Wynne H, Wood P et al. Dietary vitamin K influences intra-individual variability in anticoagulant response to warfarin. Br J Haematol, 2004;124:348–354. 4. Pedersen FM, Hamberg O, Hess K, Ovesen L. The effect of dietary vitamin K on warfarin-induced anticoagulation. J Intern Med, 1991;229:517–520. 5. Ansell J, Hirsh J, Poller L, Bussey H, Jacobson A, Hylek E. The pharmacology and management of the vitamin K antagonists: the

6.

7.

8.

9.

Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest, 2004;126:204S–233S. Franco V, Polanczyk CA, Clausell N, Rohde LE. Role of dietary vitamin K intake in chronic oral anticoagulation: prospective evidence from observational and randomized protocols. Am J Med, 2004;116:651–656. Nutescu EA, Shapiro NL, Ibrahim S, West P. Warfarin and its interactions with foods, herbs and other dietary supplements. Expert Opin Drug Saf, 2006;5:433–451. Schurgers LJ, Shearer MJ, Hamulyak K, Stocklin E, Vermeer C. Effect of vitamin K intake on the stability of oral anticoagulant treatment: dose-response relationships in healthy subjects. Blood, 2004;104:2682–2689. Karlson B, Leijd B, Hellstrom K. On the influence of vitamin K-rich vegetables and

© 2013 John Wiley & Sons Ltd

wine on the effectiveness of warfarin treatment. Acta Med Scand, 1986;220:347– 350. 10. de Assis MC, Rabelo ER, Avila CW, Polanczyk CA, Rohde LE. Improved oral anticoagulation after a dietary vitamin k-guided strategy: a randomized controlled trial. Circulation, 2009;120:1115–1122. 3 p following 22. 11. Huang IT. Constructing a provisional Taiwan food phylloquinone content database and its application. In: Constructing a provisional Taiwan food phylloquinone content database and its application, Master degree. New Taipei City: Fu Jen Catholic University, 2007, 138. 12. U.S. Department of Agriculture ARS. Vitamin K (phylloquinone) (mcg) Content of Selected Foods per Common Measure, Sorted by Nutrient Content. In: Vitamin K

Journal of Clinical Pharmacy and Therapeutics, 2014, 39, 56–60 59

C.-H. Chang et al.

Interaction of dietary vitamin K

(phylloquinone) (mcg) content of selected foods per common measure, sorted by nutrient content, 2010. Available at: http:// www.ars.usda.gov/SP2UserFiles/Place/ 12354500/Data/SR23/nutrlist/sr23w430. pdf (accessed 28 November 2010). 13. Kamao M, Suhara Y, Tsugawa N et al. Vitamin K content of foods and dietary vitamin K intake in Japanese young women. J Nutr Sci Vitaminol (Tokyo), 2007;53:464– 470. 14. Ministry of education, culture, sports, science and technology Japan. The Standardized Ingredient for Japanese Food in 2010. Available at: http://www.mext.go.jp/b_menu/shingi/gijyutu/gijyutu3/attach/1299208.htm (accessed 28 February 2013). 15. Sconce E, Khan T, Mason J, Noble F, Wynne H, Kamali F. Patients with unstable control have a poorer dietary intake of vitamin K

compared to patients with stable control of anticoagulation. Thromb Haemost, 2005;93: 872–875. 16. Kurnik D, Loebstein R, Rabinovitz H, Austerweil N, Halkin H, Almog S. Over-thecounter vitamin K1-containing multivitamin supplements disrupt warfarin anticoagulation in vitamin K1-depleted patients. A prospective, controlled trial. Thromb Haemost, 2004;92:1018–1024. 17. Trumbo P, Yates AA, Schlicker S, Poos M. Dietary reference intakes: vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, molybdenum, nickel, silicon, vanadium, and zinc. J Am Diet Assoc, 2001;101:294–301. 18. Garber AK, Binkley NC, Krueger DC, Suttie JW. Comparison of phylloquinone bioavailability from food sources or a supplement in human subjects. J Nutr, 1999;129:1201–1203.

© 2013 John Wiley & Sons Ltd

19. Gijsbers BL, Jie KS, Vermeer C. Effect of food composition on vitamin K absorption in human volunteers. Br J Nutr, 1996;76: 223–229. 20. Warren Grant Magnuson Clinical Center. National Institutes of Health Drug Interaction Task Force. Important Information to Know When You are Taking: Coumadinâ and Vitamin K. 2003. . Available at: http://ods.od.nih. gov/pubs/factsheets/coumadin1.pdf (accessed 30 September 2012). 21. Damon M, Zhang NZ, Haytowitz DB, Booth SL. Phylloquinone (vitamin K1) content of vegetables. J Food Compost Anal, 2005;18:751–758. 22. Ferland G, Sadowski JA. Vitamin K1 (phylloquinone) content of green vegetables: effects of plant maturation and geographical growth location. J Agric Food Chem, 1992;40:1874–1877.

Journal of Clinical Pharmacy and Therapeutics, 2014, 39, 56–60 60