See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/7978184

Decreased Warfarin Effect After Initiation of High-Protein, Low-Carbohydrate

Diets

Article  in  Annals of Pharmacotherapy · May 2005

DOI: 10.1345/aph.1E454 · Source: PubMed

CITATIONS READS
21 1,050

3 authors:

Stuart J Beatty Bella H Mehta

The Ohio State University The Ohio State University
14 PUBLICATIONS   251 CITATIONS    37 PUBLICATIONS   289 CITATIONS   

SEE PROFILE SEE PROFILE

Jennifer Rodis
The Ohio State University
44 PUBLICATIONS   301 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Partner for Promotion View project

All content following this page was uploaded by Bella H Mehta on 23 May 2014.

The user has requested enhancement of the downloaded file.

Decreased Warfarin Effect After Initiation of High-Protein,
Low-Carbohydrate Diets

Stuart J Beatty, Bella H Mehta, and Jennifer L Rodis

OBJECTIVE: To report 2 cases of decreased international normalized ratio (INR) after initiation of a high-protein, low-carbohydrate diet.

CASE SUMMARIES: Case 1. A 67-year-old white woman had been receiving warfarin for 3 years for venous thromboembolism. After
initiation of a high-protein, low-carbohydrate diet, the patient required a 22.2% increase (from 45 to 57.5 mg/wk) in warfarin dose.
Her INR remained in the therapeutic range on this dose for 8 weeks. When the patient stopped the high-protein, low-carbohydrate
diet, a decrease back to the original warfarin dose was required to return to a therapeutic INR. Case 2. A 58-year-old white man had
been receiving warfarin for 8 years for a cerebrovascular accident. Initiation of a high-protein, low-carbohydrate diet resulted in a
30% increase (from 26.25 to 37.5 mg/wk) in warfarin dose. His warfarin dose was reduced to the original dose after he stopped the
high-protein, low-carbohydrate diet.
DISCUSSION: The Naranjo probability scale indicated a possible adverse effect between warfarin and high-protein diets. High-protein
diets have been shown to increase serum albumin levels. This may result in more warfarin binding to serum albumin, thereby
decreasing the anticoagulant effect of warfarin. The increase of albumin occurs rapidly after initiation of a high-protein diet and
appears to promptly affect anticoagulation therapy with warfarin.
CONCLUSIONS: These cases indicate a significant interaction between high-protein, low-carbohydrate diets and warfarin therapy.
Patients receiving warfarin therapy should be educated on and monitored for the potential interaction that occurs with warfarin
therapy and high-protein, low-carbohydrate diets.
KEY WORDS: high-protein, low-carbohydrate diet; warfarin.

Ann Pharmacother 2005;39:744-7.

Published Online, 8 Mar 2005, www.theannals.com, DOI 10.1345/aph.1E454

t is well documented that warfarin sodium interacts with
Iincrease
multiple drugs, often resulting in a clinically significant
or decrease in the international normalized ratio
(INR). Dietary foods high in vitamin K are also known to
interact with warfarin and result in a clinically significant
decrease in the INR. Due to these known interactions, pa-
tients receiving warfarin are instructed to report any changes
in other medication therapy to a healthcare provider and
maintain a consistent dietary intake of vitamin K–contain-
ing foods to prevent INR fluctuation.1 Aside from the inter-
action between vitamin K–containing foods and warfarin,
there is minimal information available on other potential in-
teractions between warfarin and dietary intake.
High-protein, low-carbohydrate diets, such as the Atkins
Diet and South Beach Diet, have become popular in the US
in recent years. These diets instruct patients to increase di-
etary intake of protein and to reduce carbohydrate-contain-
ing foods. Both diets include an induction phase, in which
individuals are to eliminate almost all carbohydrates from
their diet for a minimum of 2 weeks.2,3
There is limited information available about the effect
of high-protein, low-carbohydrate diets on various phar-
macologic treatments. More specifically, as of August
2004, no information was found on the effects these diets
may have on the INR of patients taking warfarin. We de-
scribe 2 patients who experienced clinically significant re-
ductions in INR shortly after the initiation of high-protein,
low-carbohydrate diets. Use of the Naranjo probability
scale4 indicated a possible adverse effect between high-
protein, low-carbohydrate diets and warfarin therapy in the
reported patients. Written informed consent was provided
by both patients.
Case Reports
CASE 1

A 67-year-old white woman had been receiving care in our pharma-

cist-managed clinic for approximately 3 years for the management of
Author information provided at the end of the text. warfarin therapy. She was prescribed warfarin for pulmonary embolism

744 ■ The Annals of Pharmacotherapy ■ 2005 April, Volume 39 www.theannals.com

Downloaded from aop.sagepub.com by guest on October 11, 2013
and deep-vein thrombosis. Her past medical history was also significant
for non-Hodgkin’s lymphoma, which had been in remission for 5 years,
Table 1. Anticoagulation Therapy History of Patient 1
hypothyroidism, osteoarthritis, and chronic back pain. Besides warfarin, Dose New Dose
other medications reported by the patient included levothyroxine 100 Date (mg/wk)a INR (mg/wk) Comments
µg/day, sertraline 25 mg/day, potassium chloride 20 mEq/day, acetamin-
ophen as needed, and oxycodone as needed. 6/2/03 45 2.5 same taking warfarin tablets
from Barr Laboratories
From May 2002 to June 2003, the patient’s INR remained within the
therapeutic range (2.0–3.0) at 11 of 14 visits, which was maintained with 6/30 45 1.4 1 dose of 10, started Atkins diet in the
then 45 mg past 2–4 wk
warfarin 45 mg/wk. She experienced 2 subtherapeutic INRs (1.5, 1.8)
and one supratherapeutic INR (3.2) during this time. Each of these val- 7/14 45 1.7 1 dose of 10,
then 45 mg
ues returned within the therapeutic range at the next visit and did not re-
quire maintenance dose adjustment. 7/23 45 1.5 1 dose of 10, taking warfarin tablets
At a scheduled visit on June 30, 2003, an INR value of 1.4 showed a then 50 mg from Geneva Pharma-
ceuticals for the past
significant drop from the previous INR reading of 2.5 four weeks earlier
4 wk
(Table 1). The patient reported the only change in her lifestyle since the last
8/6 50 1.2 1 dose of 10, instructed to switch back
appointment was initiation of the Atkins Diet approximately 3 weeks earli-
then 50 mg to warfarin tablets from
er. Over the next 12 weeks, she had subtherapeutic INR readings at all of
Barr Laboratories and
her 8 visits. She remained on the Atkins Diet throughout this time period. use no other manu-
Notably, the patient was taking warfarin from 2 different manufactur- facturer
ers during this time: Barr Laboratories prior to July and, in July, Geneva 8/11 50 1.9 1 dose of 7.5,
Pharmaceuticals. She was instructed to switch back to the warfarin then 50 mg
tablets from Barr Laboratories on August 6; she has continued to use 8/25 50 1.6 1 dose of 10, in emergency depart-
warfarin tablets from this manufacturer since that visit. then 50 mg ment for leg pain (INR
Nearly 12 weeks after the patient started on the Atkins diet, she regained 2.2); missed dose on
a therapeutic INR with a maintenance dose of 57.5 mg/wk. This was an 8/18; deep venous
overall 22.2% increase in the dose after she initiated the high-protein, low- thrombosis ruled out
carbohydrate diet. The patient continued the Atkins diet, and the INR was 9/3 50 1.7 1 dose of 10,
within the therapeutic range at all visits over the next 8 weeks. The mainte- then 57.5 mg
nance dose of 57.5 mg/wk was continued throughout this time. 9/11 57.5 1.9 1 dose of 10,
On November 20, the patient had a supratherapeutic INR value de- then 57.5 mg
spite reporting that she had missed a warfarin dose the previous week. 9/18 57.5 2.4 same
She also said that she had taken a short course of prednisone during this 10/2 57.5 2.0 same
time and had stopped the Atkins diet. Prednisone is known to cause an 10/30 57.5 2.2 same
increase in INR when used concomitantly with warfarin.1 No other sig-
11/20 57.5 4.0 hold 1 dose, then may have missed 1
nificant changes were reported. The INR remained supratherapeutic at 2 57.5 dose; no longer
consecutive visits over the next 2 weeks despite reductions in the war- on Atkins diet
farin dose. The dose was adjusted 3 more times over a 2-week period un- 11/26 57.5 4.2 hold 1 dose, then has been taking pred-
til the INR returned to therapeutic range on December 22 at a mainte- take 1 dose of nisone therapy and is
nance warfarin dose of 40 mg/wk. 2.5 mg, then 57.5 to take for 2 more days;
Since the return of the INR to the therapeutic range, the patient’s war- no longer on Atkins
farin dose has been adjusted 3 times in 9 visits over a 5-month period. The diet and has been
INR is currently well controlled with a warfarin dose of 45 mg/wk, which drinking Slim Fast
is the same dose she was taking prior to initiation of the Atkins diet. 12/1 57.5 4.4 hold 1 dose, then last dose of prednisone
take 2 doses of 11/25
2.5 mg
CASE 2 12/4 1.4 1 dose of 7.5 mg,
then 1 dose of
A 58-year-old white man was receiving anticoagulation management 5 mg, then 2
in our pharmacist-managed clinic for 8 years after a cerebrovascular acci- doses of 7.5 mg
dent and septal defect diagnoses. His past medical history was also signif- 12/8 2.7 45
icant for hypertension, gastroesophageal reflux disease, benign prostatic
12/15 45 3.3 40
hypertrophy, and depression. Medications in addition to warfarin included
12/22 40 3.0 same thyroid medication de-
amlodipine 5 mg/day, sertraline 50 mg/day, esomeprazole 20 mg/day,
creased since last visit,
simvastatin 20 mg/day, finasteride 5 mg/day, and doxazosin 8 mg/day.
but pt. unsure of dose
From May 2003 to January 2004, the patient’s INR remained within
12/29 40 3.2 37.5 moving to Florida for the
the therapeutic range of 2.0–3.0 at 7 of 8 visits with a warfarin dose of
winter
26.25 mg/wk. At the clinic visit on January 22, 2004, his INR was at a
1/20/04 37.5 1.5 42.5 reports eating more
subtherapeutic level of 1.5 (Table 2). He reported starting the South
vegetables recently
Beach Diet approximately 3 weeks earlier. His warfarin dose was in-
1/30 42.5 1.9 same still not on Atkins diet
creased at 2 consecutive visits over 2 weeks until the INR was within the
2/11 42.5 2.7 same
therapeutic range. The INR returned to the therapeutic range after the
maintenance dose of warfarin was increased by 30% (from 26.25 to 37.5 3/10 42.5 2.0 same
mg/wk) since initiation of the high-protein diet. 4/12 42.5 2.0 45
The INR remained in the therapeutic range with that dose for 2 con- 4/26 45 3.1 same
secutive visits. On March 31, the patient had a supratherapeutic INR of 5/6 45 2.6 same
3.6 without an identifiable cause, followed by 2 more supratherapeutic 5/27 45 2.4 same Synthroid dose in-
INR values, which resulted in decreases in the maintenance dose. The creased from 0.1 to
patient reported during this time that he had been introducing carbohy- 0.125 mg/day since
drates back into the diet and was no longer following the South Beach last visit
Diet. By this time, the maintenance dose had been decreased and the
INR = international normalized ratio.
INR had returned to the therapeutic range. The patient was maintained a
Weekly sum of the daily maintenance dose; does not include bolus
on a warfarin dose of 26.25 mg/wk for 3 of 3 visits over 5 weeks, which or withheld doses.
was the same dose used prior to initiation of the South Beach Diet.

www.theannals.com The Annals of Pharmacotherapy ■ 2005 April, Volume 39 ■ 745

SJ Beatty et al.

Discussion viduals had a statistically significant increase (p < 0.001) in

albumin and total protein concentrations while consuming
An increasing number of Americans are attempting to a high-protein diet compared with a high-carbohydrate
lose weight by incorporating a high-protein, low-carbohy- diet. This increase was determined to be directly related to
drate diet into their lifestyle. These diets typically alter the the percentage of dietary intake of protein.
dietary intake of individuals, but they do not necessarily Warfarin is a medication that is highly protein bound.1
result in increased amounts of vitamin K–containing foods. An increase in albumin levels appears to occur within 10
Neither of the patients discussed here reported any change days of patients increasing their intake of dietary protein.6
in their vitamin K intake after initiation of the high-protein This increase in albumin may cause increased binding of
diets. Nonetheless, one limitation to the case report is that warfarin to albumin, resulting in less warfarin available for
dietary intake is based on self-reporting, making it difficult anticoagulant effects. This would explain why both of our
to verify the intake of vitamin K–containing foods and the patients experienced a sudden decrease in INR shortly af-
potential impact on INR values. ter initiation of a high-protein, low-carbohydrate diet.
It is listed in the warfarin package insert that numerous In a study conducted by Fagan et al.,7 patients received
factors, including changes in diet and physical state, may af- single doses of propranolol and theophylline at different
fect the anticoagulant effect of warfarin.1 Despite this listing, times to assess the relationship between dietary intake and
no information was discovered in the literature, as of August drug clearance. Patients had a 32% higher clearance of the-
2004, reporting the effects of weight loss on INR values. ophylline while on a high-protein diet compared with a
A small study described a decrease in the INR after pa- high-carbohydrate diet and a 74% higher clearance of pro-
tients increased their walking time; however, the study did pranolol while on a high-protein diet compared with a high-
not measure or discuss weight loss as a factor for de- carbohydrate diet. The authors of this study proposed that
creased INR values.5 Neither of our patients reported any an induction of the cytochrome P450 system as a result of a
increase in physical activity to cause suspicion of decreased high-protein diet is the cause of the altered clearance rates.
INR by this mechanism. Warfarin, propranolol, and theophylline are all metabolized
Both of the patients were adherent and consistently by the CYP450 isoenzyme system; however, all 3 medica-
within their therapeutic INR range prior to the initiation of tions use different isoenzymes as their major metabolic
high-protein, low-carbohydrate diets. Both patients experi- pathway. Additional studies are needed to assess the effect
enced a rapid, significant drop in INR values after begin- of dietary intake of protein on the entire CYP450 system.
ning the diets. A proposed mechanism for a decreased INR This study shows evidence of altered drug clearance by a
in patients following a high-protein, low-carbohydrate diet high-protein diet. The effects of a high-protein diet on pro-
involves the effect of protein intake on albumin. A small pranolol and theophylline clearance may be similar to the way
study evaluated the effects of dietary intake on various lab- a high-protein diet affects warfarin. Similar to warfarin, pro-
oratory values in 6 healthy patients. Every patient received pranolol is also a highly protein-bound drug, while theophyl-
each of the following 3 controlled diets: high protein, high line is only moderately protein bound.8,9 It is possible that this
carbohydrate, and high fat for 10 days per diet.6 The indi- difference in protein binding is the reason for a larger effect
on the clearance rate of propranolol than of theophylline.
A number of studies have looked at INR fluctuation upon
Table 2. Anticoagulation History of Patient 2
substitution of different warfarin tablet manufacturers.10-12
Case reports have supported our statement that changing
Dose New Dose warfarin manufacturers may alter INR values.9 Conversely,
Date (mg/wk)a INR (mg/wk) Comments
one clinical study showed no significant change in INR val-
12/22/03 26.25 2.5 same
ues regarding change of manufacturers,11 while another in-
1/22/04 26.25 1.5 1 dose of 7.5, started South Beach
then 31.25 mg Diet 3 wk ago vestigation showed a statistically, but not clinically, signifi-
1/28 31.25 1.5 1 dose of 7.5, still on South Beach cant change in INR values.12 Patient 1 began taking warfarin
then 35 mg Diet from a different manufacturer around the same time as the
2/5 35 2.1 37.5
initiation of the high-protein diet. However, when she was
2/19 37.5 2.6 same
3/11 37.5 3.0 same still on South Beach
switched back to warfarin tablets from the original manu-
Diet facturer within a month, the INR remained subtherapeutic.
3/31 37.5 3.6 1 dose of 2.5, In addition, the patient’s maintenance dose was returned to
then 37.5 mg
the same dose she was taking before initiation of the high-
4/26 37.5 3.3 1 dose of 2.5, stopped South Beach
then 35 mg Diet protein, low-carbohydrate diet. Based upon these facts, it is
5/13 35 4.3 hold 1 dose, then highly unlikely that the manufacturer substitution was the
31.25 mg cause of the decreased INR in this case.
5/20 31.25 2.6 26.25
6/3 26.25 2.0 same
Conclusions
INR = international normalized ratio.
a
Weekly sum of the daily maintenance dose; does not include bolus The potential interaction between warfarin and high-
or withheld doses.
protein, low-carbohydrate diets is significant, especially as

746 ■ The Annals of Pharmacotherapy ■ 2005 April, Volume 39 www.theannals.com


more people turn to these diets in attempts to lose weight.
The resulting decrease in the INR appears to occur rapidly
after initiation of the diet, placing patients on warfarin thera-
py at a high risk of thromboembolic events and potentially
death. Also, it appears an increase in the INR occurs rapidly
after discontinuation of the high-protein, low-carbohydrate
diet, placing patients at increased risk of major and minor
hemorrhagic events. Therefore, we recommend that patients
currently receiving warfarin therapy and planning to start or
stop a high-protein diet should inform their healthcare
provider, and more frequent INR monitoring should occur.
The effect of high-protein, low-carbohydrate diets on
pharmacologic therapy needs to be further studied. In addi-
tion to the effects mentioned in these case reports, there is ev-
idence of increased clearance of other drugs during high in-
take of dietary protein, making interactions with other medi-
cations and high-protein diets possible. Until more research is
performed on the effects of medications and high-protein,
low-carbohydrate diets, it is difficult to determine how these
diets affect warfarin or any other pharmacologic agent.
Stuart J Beatty PharmD, at time of writing, Pharmacy Practice
Resident with an Emphasis in Community Care, College of Phar-
macy, The Ohio State University, Columbus, OH; now, Clinical Phar-
macist, Pharmacotherapy Clinic, MedCenter Pharmacy, Marion, OH
Bella H Mehta PharmD, Assistant Professor of Clinical Pharmacy,
Division of Pharmacy Practice and Administration, College of Phar-
macy, The Ohio State University
Jennifer L Rodis PharmD, Assistant Professor of Clinical Phar-
macy, Division of Pharmacy Practice and Administration, College of
Pharmacy, The Ohio State University
Reprints: Dr. Rodis, College of Pharmacy, The Ohio State Univer-
sity, 500 W. 12th Ave., Columbus, OH 43210-1291, fax 614/292-
1335, rodis.2@osu.edu
References
1. Package insert. Coumadin (warfarin sodium). Princeton, NJ: Bristol-My-
ers Squibb, June 2002.
2. Atkins RC. Dr. Atkins’ new diet revolution. 1st ed. New York: Harper-
Collins, 2002.
3. Agatson AS. The South Beach Diet: the delicious, doctor-designed, fool-
proof plan for fast and healthy weight loss. 1st ed. New York: Random
House, 2003.
4. Naranjo CA, Busto U, Sellers EM, Sandor P, Ruiz I, Roberts EA, et al. A
method for estimating the probability of adverse drug reactions. Clin
Pharmacol Ther 1981;30:239-45.
5. Shibata Y, Hashimoto H, Kurata C, Ohno R, Kazui T, Takinami M. In-
fluence of physical activity on warfarin therapy. Thromb Haemost
1998;80:203-4.
6. Fagan TC, Oexmann MJ. Effects of high protein, high carbohydrate, and
high fat diets on laboratory parameters. J Am Coll Nutr 1987;6:333-43.
7. Fagan TC, Walle T, Oexmann MJ, Walle UK, Vai SA, Gaffney TE. In-
creased clearance of propranolol and theophylline by high-protein com-
pared with high-carbohydrate diet. Clin Pharmacol Ther 1987;41:402-6.
8. Package insert. Inderal (propranolol hydrochloride). Philadelphia: Wyeth-
Ayerst Laboratories, November 2003.
9. Package insert. Theolair (theophylline). Northridge, CA: 3M Pharma-
ceuticals, May 1998.
10. Hope KA, Havrda DE. Subtherapeutic INR values associated with a
switch to generic warfarin. Ann Pharmacother 2001;35:183-7.
DOI 10.1345/aph.10207
11. Swenson CN, Fundak G. Observational cohort study of switching war-
farin sodium products in a managed care organization. Am J Health Syst
Pharm 2000;57:452-5.
www.theannals.com The Annals of Pharmacotherapy
View publication stats
INR Decrease with High-Protein, Low-Carbohydrate Diets
12. Witt DM, Tillman DJ, Evans CM, Plotkin TV, Sadler MA. Evaluation of
the clinical and economic impact of a brand name-to-generic warfarin
sodium conversion program. Pharmacotherapy 2003;23:360-8.
EXTRACTO
OBJETIVO: Informar 2 casos de disminución en la razón normalizada
internacional (INR, por sus siglas en inglés) después de comenzar una
dieta alta en proteína y baja en carbohidrato.
RESUMEN DEL CASO: Una mujer caucásica de 67 años en terapia con
warfarina por 3 años debido a tromboembolismo venoso requirió un
aumento de un 22.2% (de 45 a 57.5 mg/semana) en la dosis de warfarina
después de comenzar una dieta alta en proteína y baja en carbohidrato. El
INR de esta paciente se mantuvo en el rango terapéutico en la dosis
nueva por 8 semanas. Hubo que disminuir la dosis de warfarina a la dosis
original para lograr de nuevo un INR en rango terapéutico cuando la
paciente descontinuó la dieta alta en proteína y baja en carbohidrato. Un
hombre caucásico de 58 años en terapia con warfarina por 8 años debido
a un accidente cerebrovascular requirió un aumento de un 30% (de 26.25
a 37.5 mg/semana) en la dosis de warfarina. En este paciente hubo que
reducir la dosis de warfarina a la dosis original después que el paciente
descontinuó la dieta alta en proteína y baja en carbohidrato.
DISCUSIÓN: Se ha demostrado que la concentración de albúmina sérica
aumenta con dietas que son alta en proteína. Esto puede ocasionar el que
una cantidad mayor de warfarina se enlace a la albúmina sérica y por
consiguiente se disminuye el efecto anticoagulante de warfarina. El
aumento en la albúmina ocurre rápidamente luego de comenzar una
dieta alta en proteína y parece afectar prontamente la terapia
anticoagulante con warfarina.
CONCLUSIONES: Estos 2 casos son indicio de que existe una interacción
importante entre las dietas que son alta en proteína y baja en
carbohidrato y la terapia con warfarina. Los pacientes que estén en
terapia con warfarina deben recibir educación sobre y seguimiento para
la interacción potencial entre la terapia con warfarina y las dietas que
son alta en proteína y baja en carbohidrato.
Luz M Gutiérrez
RÉSUMÉ
OBJECTIF: Rapporter 2 cas de diminution du rapport normalisé
international (RNI) survenus suite à l’initiation d’une diète riche en
protéines et pauvre en glucides.
RÉSUMÉ: Cas #1: Une femme de 67 ans prenait de la warfarine depuis 3
ans suite à une thrombo-embolie veineuse. Après avoir débuté une diète
riche en protéines et pauvre en glucides, la dose de warfarine a dû être
augmentée de 22.2% (45 à 57.5 mg/sem). Suite à cet ajustement de
dose, le RNI est demeuré dans l’écart thérapeutique pendant 8 semaines.
Lorsque la patiente cessa cette diète, un retour à la dose initialement
prescrite a été nécessaire pour maintenir le RNI dans l’écart
thérapeutique. Cas #2: Un homme de 58 ans prenait de la warfarine
depuis 8 ans suite à un accident vasculaire cérébral. Après avoir débuté
une diète riche en protéines et pauvre en glucides, la dose de warfarine a
dû être majorée de 30% (26.25 à 37.5 mg/sem). Après avoir cessé cette
diète, la dose de warfarine a été réduite à la dose initialement prescrite.
DISCUSSION: Les diètes riches en protéines augmentent le taux
d’albumine sérique. La warfarine se lie alors davantage à l’albumine, ce
qui résulte en une diminution de l’effet anticoagulant de la warfarine.
L’augmentation du taux d’albumine survient tôt après le début d’une
diète riche en protéines et influence rapidement l’efficacité de
l’anticoagulothérapie à la warfarine.
CONCLUSIONS: Ces rapports de cas indiquent qu’il existe une interaction
significative entre la diète riche en protéines et pauvre en glucides et la
warfarine. Les patients qui reçoivent de la warfarine devraient être
informés de cette interaction potentielle et faire l’objet d’un suivi
approprié.
Alain Marcotte
■ 2005 April, Volume 39 ■ 747