Acta Med Scand 1986; 220: 347-50

On the Influence of Vitamin K-rich Vegetables and Wine

on the Effectiveness of Warfarin Treatment

B. KARLSON, B. LEIJD and K. HELLSTROM

From the Department of Internal Medicine, St Erik's Hospital, Stockholm, Sweden

ABSTRACT. Karlson B, Leijd B, Hellstrom K (Department of Internal Medicine, St

Erik's Hospital, Stockholm, Sweden.) On the influence of vitamin K-rich vegetables and
wine on the effectiveness of warfarin treatment. Acta Med Scand 1986; 220: 347-50.
Thrombotest (TT) values were studied in patients fed an ordinary diet and on continuous
and well controlled warfarin therapy because of deep venous thrombosis or pulmonary
embolism. The aim was to characterize the effect of single and multiple administrations
(daily during one week) of vitamin K1 (Konakion"), vitamin K-rich vegetables such as
spinach and broccoli, and table wine. Single administration of 250 pg vitamin K,, 250 g
spinach, 250 g broccoli and 37.5 cl wine did not result in TT-values outside the therapeutic
range. However, when Konakion", broccoli and spinach were given daily during one week
the 'IT-values tended to rise above the therapeutic limit, requiring dose adjustment. On the
basis of this study it appears that excessive intake of vitamin K-rich food and a moderate
intake of alcohol on one occasion may be permitted during anticoagulant therapy. Key
words: vegetables, vitamin K , alcohol, warfarin.

In the management of anticoagulant therapy it is necessary to balance between preventing

thrombosis and avoiding hemorrhagic complications. The prescription of a fixed dose of
anticoagulants in long-term therapy is supposed to keep patients in a steady state with
regard to factors involved in the clotting mechanisms. However, as coumarines such as
warfarin interact with the inactive precursors to the vitamin K dependent proteins (l), it is
conceivable that irregular intakes of vitamin K modify the effect of therapy. The doses
required may vary from patient to patient. Another variable worth observing is alcohol,
the intake of which is reported to increase as well as decrease the levels of coagulation
factors (2, 3). Moreover, as alcohol may stimulate the activity of hepatic enzymes
responsible for inactivation of coumarines (4), the overall consequence of alcohol con-
sumption may be variable and unpredictable.
Although anticoagulants are commonly used in routine care, there seems to be a lack of
studies indicating how to advise patients with regard to intakes of vitamin K-rich food and
alcohol. The present investigation was undertaken to evaluate the influence of some
foodstuffs known to be rich in vitamin K (broccoli and spinach), and of wine on warfarin
therapy.

STUDY POPULATION AND METHODS

Patients
Altogether 10 women and I 1 men (48-82 years old) took part in the study. Before the start of warfarin
(Waran") therapy the subjects had been advised to keep to their usual diet and to avoid excessive
intakes of green vegetables, cabbage, liver and alcohol. They had all been on continuous treatment for
at least 3 months because of deep venous thrombosis or pulmonary embolism. Therapy was moni-
tored with Thrombotest (TT)as described by Owren (5). During this treatment the TT-values were
reasonably constant and within the therapeutic range (5-15%). All other medication was kept
constant in the 10 patients treated with digoxin, furosemide, glibenclamide andlor spironolactone.

Abbreviation: 'IT-values =thrombotest values.

348 B . Karlson et al. Acta Med Scand 1986; 220

I
c 37.6 ci IaMO WIN) 35 35 *-• spinach
e-.. 2x) pg vitamin K, Vitamin K ,

. . . 7d.W
- ....... *.?..9 broccoli

5 f ? .
0 1 2
,
3
,
4 6 6
30 30.
.. ,
TT % 25 25.
15

20 20.

10
15 15-

.......
c. spinach
bmccdi
1c 10-

5 4 5 5. + + b b O + + l
P
0 1 2 3 4 5 6 7dap 0 1-2 3-4 5-6 7 - 8 day3

Fig. 1 Fig. 2.
Fig. 1 . TT-values after single intakes (arrow) of 250 pg vitamin K,, 37.5 cl table wine, 250 g spinach or
250 g broccoli. * pCO.05.
Fig. 2. TT-values during daily intakes of 100, 250 or 250 pg vitamin K,, 250 g spinach or 250 g
broccoli. * p<0.05, ** p<O.OI, *** p<O.Ool.

When warfarin therapy was to be withdrawn the patients were invited to participate in the current
investigation. Those who were interested gave their informed consent. The study was approved by
the ethical committee at Karolinska Institutet, Stockholm.
Protocol
Ten subjects were given single doses of 250 pg Konakion” (vitamin K I , Phytomenadione), 250 g
broccoli, 250 g spinach or 37.5 cl table wine (41 g alcohol). Three of them participated together with 8
other subjects in a second set of experiments. These 11 subjects were given either 250 g broccoli, 250
g spinach, 100, 250 or 500 pg Konakion per day during one week. Thrombotest was followed almost
daily. The patients continued with their regular doses of warfarin during the whole study. Before the
start of a new experiment the TT-value was permitted to return to the initial level, which in all
instances was within the therapeutic limits.
Material and Meihods
Broccoli and spinach, purchased deep-frozen from Foodia (KF:s Livsmedelsindustri, Staffanstorp,
Sweden), were heated in Lightly salted water. The content of vitamin K, was calculated to be in the
range 160-500 pg in 250 g broccoli and 300-800 pg in 250 g spinach (6). The Thrombotest reagent was
obtained from Nyegaard A-G.
Statistics
Data are presented as mean k SEM. Wilcoxon’s rank test was used in the calculation of differences.

RESULTS
A single dose of 250 vg Konakionm as well as intake of 250 g broccoli or spinach tended to
elicit a slight rise of the ‘IT-values. In contrast, 37.5 cl wine had no effect (Fig. 1). Daily
administration of 100 pg Konakion had only a small effect on the anticoagulation therapy.
Acta Med Scand 1986; 220 Vegetables and wine during warfarin 349

Following intakes of 250 or 500 pg vitamin K1 per day, the TT-values exceeded the
therapeutic limit within a few days (Fig. 2). The corresponding rise after ingestion of 250 g
broccoli per day was similar to that observed with a daily dose of 250 pg Konakion@.The
response in TT-values after intakes of 250 g spinach tended to be even more marked.

DISCUSSION
Studies of the metabolism of vitamin K have been hampered by technical difficulties due
to the small levels of the vitamin in food, blood and tissues. Until recently, bio-assay was
the only method available for determining vitamin K in food. The results obtained have not
been consistent (6, 7).
It appears that the intake of vitamin K in the “normally mixed diet” in Western
countries varies between 300 and 500 &day. If so, the doses of Konakion administered in
the current study (100-500 pg/day) should represent a significant elevation of the vitamin
K intake. However, judging from the current data, the clotting mechanisms are rather
insensitive to short-term variations in vitamin K1 intake as a single dose of 250 pg
Konakion@had insignificant effects on the TT-values.
The absorption of vitamin K1 is reported to be very effective (8). Considering such
findings, the calculated intake (SOrr800 pg) after ingestion of spinach and broccoli seemed
to be overestimated in these vegetables, which are known to be rich in vitamin K1. To
what extent the discrepancies are due to a low vitamin content in the vegetables used here
or to changes during storage and preparation is unknown (9).
The current study indicates that single intakes of vitamin K1-rich vegetables have no
major effect on the TT-values. However, in keeping with some case reports dealing with
warfarin resistance (10-12), a continuous intake of high amounts of spinach or broccoli
may interfere with warfarin therapy to an extent requiring dose adjustments. Similar
effects have been reported in patients with liquid nutrition receiving unrecognized supple-
mentation of vitamin K (13). Vitamin K2 synthesized by intestinal microorganisms is
another source of vitamin K. The size and regulation of vitamin K2 production and thus its
importance to blood coagulation are almost unknown (8).
As indicated by the current and some previous reports (14, 15), there is no good reason
to forbid occasional intakes of moderate amounts of wine by warfarin-treated patients with
a normal liver function. The effects on clotting mechanisms of more regular drinking, even
in moderate amounts, remain unclear.

REFERENCES
1. Stenflo J, Fennland P, Egan W, Roepstorff P. Vitamin K dependent modifications of glutamic
acid residues in prothrombin. Proc Mol Acad Sci 1974; 71: 2730-3.
2. Wanis R. Effect of ethyl alcohol on some coagulation factors in man during anticoagulant
therapy. Ann Med Exp Fenn 1963; 41: 45-54.
3. Riedler G . Einfluss der Alcohols auf die Anticoagulantien Therapie. Thromb Diath Haemorrh
1966; 16:613-35.
4. Kater RMH, Roggin G , Tobon F, Zieve P, Iber FL. Increased rate of clearance of drugs from the
circulation of alcoholics. Am J Med Sci 1969; 258: 35-9.
5. Owren PA. Thrombotest. A new method for controlling anticoagulant therapy. Lancet 1959;
2: 754-8.
6. Parrish D. Determination of vitamin K in food. A review. Food Science and Nutrition 1980;
337: 352.
7. Olson RE. Vitamin K. In: Goodhart RS, Skils ME, eds. Modem nutrition in health and disease.
Philadelphia: Les H Febiger, 1980; 170-80.
8. Barkham P, Shearer MJ. Metabolism of vitamin K. Proc Roy SOCMed 1977; 80: 93-6.
350 B . Karlson et al. Acta Med Scand 1986; 220

9. Richardson LR, Wilkes S, Ritchey SJ. Comparative vitamin K activity of frozen irradiated and
heat-processed foods. J Nutr 1961; 73: 369-73.
10. Quich A. Leafy vegetables in diet after prothrombin time in patients taking anticoagulant drugs.
JAMA 1969; 187:27.
11. Udall JA Kroch LB. A modified method of anticoagulant therapy. Curr Ther Res 1968;
10:207-1 1.
12. Kempin S. Warfarin resistance caused by broccoli. Letter to the editor. N Engl J Med 1983;
308: 1229.
13. O’Reilly R, Rytand D. Resistance to warfarin due to unrecognized vitamin K supplementation. N
Engl J Med 1980; 303: 1661.
14. O’Reilly R. Lack of effect of mealtime wine on the hypoprothrombinernia of oral anticoagulants.
Am J Med Sci 1979; 277: 189-94.
15. O’Reilly R. Lack of effect of fortified wine ingested during fasting and anticoagulant therapy.
Arch Intern Med 1981; 141:45&9.

Received Dec. 2, 1985.

Correspondence: Bo Karlson, Reimersholmsgatan 75, S-117 40 Stockholm, Sweden.