SEARCH

SEARCH BY CITATION

Keywords:

  • fasting;
  • humans;
  • international normalized ratio;
  • Muslim;
  • warfarin

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References

Background

Anticoagulation with warfarin is influenced by dietary changes but the effect of fasting on warfarin therapy is unknown.

Objectives

To study changes in international normalized ratio (INR) and the percentage of time within therapeutic range (%TTR) before, during and after the Muslim fasting month (Ramadan) in stable warfarinised Muslim patients.

Methods/Patients

In this prospective study, weekly INR readings were taken at home visits from participating patients during three study periods: before, during and after Ramadan. Readings were blinded to patients and their primary physicians except for when pre-set study endpoints were reached.

Results

Among 32 participating patients, mean INR increased by 0.23 (P = 0.006) during Ramadan from the pre-Ramadan month and decreased by 0.28 (P < 0.001) after Ramadan. There was no significant difference (P = 1.000) in mean INR between the non-Ramadan months. %TTR declined from 80.99% before Ramadan to 69.56% during Ramadan (P = 0.453). The first out-of-range INR was seen around 12.1 days (95% CI, 9.0–15.1) after the start of fasting and returned to range at about 10.8 days (95% CI, 7.9–13.7) after Ramadan. Time above range increased from 10.80% pre-Ramadan to 29.87% during Ramadan (P = 0.027), while time below range increased from 0.57% during Ramadan to 15.49% post-Ramadan (P = 0.006). No bleeding or thrombotic events were recorded.

Conclusions

Fasting significantly increases the mean INR of medically stable patients taking warfarin and the likelihood of having an INR above therapeutic targets. For patients maintained at the higher end of INR target ranges or at increased risk of bleeding, closer monitoring or dosage adjustment may be necessary during fasting.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References

Oral vitamin K antagonist (VKA) has been the mainstay of anticoagulation therapy for more than half a century, with warfarin as the most commonly used compound [1]. It has established efficacy and safety in primary and secondary prevention of thromboembolic events [2]. However, due to its small therapeutic window and its potential interactions with food and medications, regular monitoring and dose adjustments are required. The potential influence of a person's dietary pattern on VKA's therapeutic effect has raised the question of whether intermittent, periodic fasting, as practiced in some religions, will have an effect on an individual's anticoagulation therapy [3].

During the ninth month of the Muslim calendar or Ramadan, Muslims fast from sunrise to sunset. Most Muslim patients taking warfarin observe Ramadan and alter the timing of taking their medicines. The effects of fasting coupled with changes in patterns of warfarin and dietary intake during Ramadan are currently not known. The current published literature linked to this subject is limited to animal models [4, 5].

To determine its potential effects on patients, we conducted a prospective observational study on a cohort of Muslim patients taking warfarin who fasted during the holy month of Ramadan.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References

The study was conducted from 20 June 2012 to 17 September 2012 with approval from our institutional review board. The database of the Anticoagulation Clinic (ACC) at Singapore General Hospital and pharmacy warfarin dispensing record were used to identify Muslim patients. Inclusion criteria for subjects were as follows: between the ages of 21 and 80 years; on steady warfarin dose with at least two consecutive measured international normalized ratios (INRs) within the target range, no less than 1 month apart preceding the study; and clinically assessed to be fit for fasting without requiring a further scheduled clinic visit during the study period. We excluded patients who were within 3 months of a thrombotic event, had unstable INR or had contraindications for fasting.

At the screening visit 2 weeks from study commencement, patients were provided with further study details and assessed by a hematologist for fitness to fast. Written informed consent was obtained from all patients prior to study enrolment. Baseline demographic information and treatment indications were recorded. Figure 1 details our patient recruitment process.

image

Figure 1. Consort study algorithm.

Download figure to PowerPoint

Study design and procedures

The study was conducted with three equal monthly observation periods – control (pre-Ramadan), intervention (Ramadan) and follow-up (post-Ramadan), during which INR measurements were taken weekly by the study team at the patients' homes. Each patient served as their own control and was required to take the prescribed warfarin dose daily and chart their intake in a dosing diary. INR readings recorded by the study team were not revealed to patients and their primary physicians. The study team was not permitted to make changes to the warfarin dose based on the weekly INR readings. Participants were also not required to visit the ACC for monitoring during the 3-month study period, except when predefined study endpoints as stated below were reached. Signs and symptoms of bleeding and thrombotic events, as well as local complications associated with finger-pricks performed for INR monitoring, were recorded.

The primary outcome measure was the deviation of the mean INR during the month of Ramadan and post-Ramadan compared with the pre-Ramadan period. Secondary outcome measures included the percentage of time spent within therapeutic range across the study periods. Clinical endpoints for termination of patients from the study included any major bleeding event and any venous or arterial thrombotic episode. Patients were asked to seek immediate medical attention should any suspicion of these clinical endpoints occur. Major bleeding was defined according to standard criteria used in clinical trials [6].

Revealing INR results to patients and their primary physician was permitted if two consecutive weekly INR readings were below 1.5 or above 4.0, or a single INR reading was above 5.0. In these instances the primary physician would be notified and the patient's participation in the study would be terminated. Apart from the above safety measures, patients could withdraw their participation at any time throughout the study. A patient's participation might also be terminated if he or she failed to comply with the prescribed warfarin dose, or failed to permit scheduled INR measurements.

All INR readings were measured using the portable Coaguchek® XS POC Coagulometer (Roche Diagnostics Asia Pacific Pte Ltd, Singapore), which has been validated in previous studies and is used for routine INR testing at our ACC [7-9].

Statistical analysis

Descriptive statistics were used to report patient demographics. The primary outcome measure was analyzed using analysis of covariance, with period of INR measurement (i.e. pre-Ramadan, Ramadan, post-Ramadan) as a fixed effect. %TTR was determined for each individual patient using Rosendaal's linear interpolation and combined to obtain the %TTR for each period [10]. One-way analysis of variance was employed to evaluate the difference in %TTR between periods. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 20.0 (IBM Corp., Armonk, NY, USA) with the a priori α set at 0.05. In addition, adherence to warfarin was estimated by the ratio of the number of days the patients consumed warfarin at the prescribed doses to the duration of study follow-up, expressed as a percentage.

Results and discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References

A total of 32 patients consented to participate in this study. Thirty patients completed the study, with two patients withdrawn after recording INR readings of more than 4.0 on two consecutive occasions during the pre-Ramadan and Ramadan periods, respectively (Fig. 1). Patient demographics, indications for warfarin and prescribed INR targets are shown in Table 1. Among this cohort, most patients suffered from venous thromboembolism (84.3%) and had an INR target of 2.0–3.0 (65.6%).

Table 1. Patient demographics
 Study population (N = 32)
Mean age, years51 7 ± 15.0
Sex (%)
Male16 (50)
Female16 (50)
Indication (%)
Venous thromboembolism (VTE)27 (84.3)
Atrial fibrillation (AF)3 (9.4)
Mechanical valve replacement (MVR)2 (6.3)
Treatment duration prescribed (%)
6 months1 (3.1)
9 months1 (3.1)
12 months0 (0.0)
Long term30 (93.8)
INR target (%)
2.0–3.021 (65.6)
2.5–3.52 (6.3)
1.5–2.54 (12.5)
2.0–2.53 (9.4)
1.5–2.01 (3.1)
1.8–2.21 (3.1)
Median number of concomitant medications (except warfarin)2 (range: 0–14)
Mean number of co-morbidities (excluding indication(s) for anticoagulation)1 (range: 0–6)

Primary and secondary outcome measures

The INR readings of the 30 patients who completed the study were used for analysis of the study endpoints. The results of the primary outcome are shown in Table 2. Mean INR readings increased by 0.23 (2.44 vs. 2.67, P = 0.006) during Ramadan when compared with pre-Ramadan. In the month following Ramadan, mean INR readings declined by 0.28 (2.67 vs. 2.38, P < 0.001) when compared with those of the Ramadan month. Mean INR readings were comparable between pre-Ramadan and post-Ramadan (2.44 vs. 2.38, P = 1.000).

Table 2. Study results: mean INR differences between study periods and percentage of time within, above and below therapeutic range
123
INR ResultsPre- RamadanRamadanMean difference P RamadanPost-RamadanMean difference P Pre- RamadanPost-RamadanMean difference P
2.44 ± 0.522.67 ± 0.57−0.2280.0062.64 ± 0.572.38 ± 0.580.283<0.0012.44 ± 0.522.38 ± 0.580.0551.000
 Pre-Ramadan (%)Ramadan (%)Post-Ramadan (%)
  1. Data are mean ± SD unless otherwise indicated.

Time within range80.9969.5669.87
Time above range10.8029.8714.64
Time below range8.220.5715.49

The overall %TTR during the 3-month study period was 73.25%. Across the three periods, the %TTR during Ramadan was lower compared with the non-Ramadan periods although this was not statistically significant (pre-Ramadan, 80.99%; Ramadan, 69.56%; post-Ramadan, 69.87%). There was, however, a 3-fold increase in the time above the therapeutic target range during Ramadan compared with pre-Ramadan (pre-Ramadan 10.80% vs. Ramadan 29.87%, P = 0.027). Conversely, during the post-Ramadan period, there was a significant rise in time below therapeutic range compared with Ramadan as patients celebrated the end of Ramadan fasting with a month-long festival (Ramadan 0.57% vs. post-Ramadan 15.49%, P = 0.006). The fluctuations in quality of anticoagulation are illustrated pictorially in Fig. 2.

image

Figure 2. Fluctuations in quality of anticoagulation across study periods.

Download figure to PowerPoint

During the study, five patients had temporary but intentional omission of warfarin (22 missed doses in total) due to menorrhagia (one patient), scheduled dental surgery (two patients) and non-anticoagulation-related sickness and hospitalization (two patients). The INR readings recorded within 5 days of the last day of these dose omissions were excluded from the analyses. In addition, there were five instances of unintentional missed doses by four patients throughout the 3-month study period. The INR readings recorded surrounding these unintentional missed doses were included in the analyses. Excluding intentional omissions, the overall adherence rate for prescribed warfarin dosing throughout the study was 99.81%.

No thrombotic or bleeding events were reported during the study period. There were also no complications arising from the weekly finger-prick monitoring of INR.

The significant increase in mean INR during Ramadan when compared with non-Ramadan periods is attributable to the effect of fasting and the consequential changes to dietary patterns and timing of food intake. Ramadan is associated with sleep deprivation as a result of long periods of praying and dehydration due to daytime withdrawal of fluids [11]. In terms of metabolic changes, it has been shown that fasting during Ramadan resulted in a temporary reduction in body weight, blood glucose and high-density lipoprotein and an increase in low-density lipoprotein levels [12]. These physical and mental stress factors during the Ramadan period are possible reasons for the increase in INR, which has been demonstrated in animal studies [5, 13]. As a result of the fewer number of meals taken each day and changes to dietary content, the amount of dietary vitamin K obtained from food may also be reduced. Multiple factors are therefore present during Ramadan that may directly or indirectly reduce the synthesis of clotting factors, resulting in the increase in INR. The return of the mean INR to baseline after Ramadan lends further credence to the suggestion that the effect of fasting on INR is not incidental. Patients usually resume their usual pattern of food intake and associated stress factors resolve post-fasting.

While statistically significant, the absolute increase in the mean INR is reassuringly small and likely to have limited clinical implications for the majority of Muslim patients. However, for patients who are maintained within a narrower INR target range because of bleeding concerns or have usual INR readings at the upper limit of the target range, such small changes in INR may result in readings above their therapeutic INR targets. The changes in the TTR patterns recorded during the study periods support this contention. An absolute TTR reduction of 11.43% (P = 0.453) was registered during Ramadan compared with the pre-Ramadan month. Importantly, the percentage of time above therapeutic range for the cohort was increased 3-fold from 10.80% to 29.87%. As bleeding risk increases exponentially with a rise in INR, this effect of fasting may increase their propensity to bleed. It has been demonstrated that an INR of more than 3.0 is associated with a 3-fold increase in risk of bleeding compared with patients who have an INR of less than or equal to 3.0 [14]. We would therefore recommend closer monitoring or pre-emptive dose adjustments in this group of patients during Ramadan. This period of close monitoring should probably extend into the post-fasting month as INR readings do not immediately return to the stable pre-fasting levels.

Following Ramadan, we observed an over-correction of INR readings, with a significant increase in percentage of time below therapeutic range from 0.57% during Ramadan to 15.49% post-Ramadan. This was also higher than the pre-Ramadan level of 8.22%. The higher proportion of time below therapeutic range could be due to increased food intake during the month-long festival after Ramadan. Along with the initial delay in return of higher INR to therapeutic range, the TTR only improved slightly from 69.56% to 69.87% (P = 1.000) during this period. Whilst the fluctuations in TTR were not statistically significant across the study periods, the clinical implications can still be important as a previous study has shown that modest changes in TTR may significantly impact clinical outcomes [15]. Clinicians may thus need to closely monitor patients who are at risk of bleeding or thrombotic complications during the post-Ramadan period.

We do, however, recognize a few limitations of this study. Potential confounders that might have impacted the INR readings included the intentional omission of warfarin due to illness or menstruation. The analyses attempted to negate the impact of this by omitting INR readings that could predictably be affected by such uncontrollable events. This study also did not evaluate the physiological effect of fasting alone. Rather it looked at the constellation of changes associated with the practice of fasting by Muslims during Ramadan in Singapore. Lastly, for ethical reasons, we excluded patients who were assessed as requiring more frequent monitoring during Ramadan. This could therefore have the impact of removing the population whose INR was most susceptible to fasting.

In conclusion, fasting during Ramadan significantly increases the mean INR of Muslim patients taking warfarin and the likelihood of having an INR above therapeutic targets. For patients maintained at the higher end of INR target ranges or at increased risk of bleeding, clinicians will need to be aware of this effect and consider more intensive monitoring or pre-emptive dosage adjustments should these patients choose to fast.

Addendum

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References

Y. F. Lai, Principal Investigator, was responsible for the following: literature search, study design, grant application, patient recruitment, data collection, data analysis, manuscript writing and manuscript reviewing. M. H. H. Cheen, Co-Investigator, was responsible for the following: study design, patient recruitment, data collection, data analysis and manuscript reviewing. S. H. Lim, Co-Investigator, was responsible for the following: study design, patient recruitment, data collection, data analysis and manuscript reviewing. F. H. I. Yeo, Co-Investigator, was responsible for the following: patient recruitment, data collection and manuscript writing. S. C. Nah, Co-Investigator, was responsible for the following: patient recruitment, data collection and manuscript reviewing. M. C. Kong, Co-Investigator, was responsible for the following: budgeting and vendor coordinating, study design and manuscript reviewing. D. Mya, Co-Investigator, was responsible for the following: patient recruitment and manuscript reviewing. L. H. Lee, Research Mentor, was responsible for the following: study design, grant application and manuscript reviewing. H. J. Ng, Research Supervisor, was responsible for the following: study design, grant application and manuscript reviewing.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References

Aung Soe Tin, Centre for Quantitative Medicine, Duke-NUS Graduate Medical School, Singapore. Chun Fan Lee, Singapore Clinical Research Institute, Singapore. patients' written informed consent for data to be published was sought prior to recruitment into the study in accordance with Singhealth Centralized Institutional Review Board's protocol.

Funding source declaration

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References

Singapore General Hospital Research Grant funded the study. The funding body neither participated in the planning and execution of the study, nor had access to study data and analyses at its conclusion.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results and discussion
  6. Addendum
  7. Acknowledgements
  8. Funding source declaration
  9. Disclosure of Conflict of Interests
  10. References
  • 1
    McArthur MC, Dzintars EK, Phillips RB, Bushardt RL. Oral anticoagulation: a review of the current and emerging therapies. JAAPA 2011; 24: 606.
  • 2
    Hirsh J, Dalen J, Anderson DR, Poller L, Bussey H, Ansell J, Deykin D. Oral anticoagulants: mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest 2001; 119: 8S21S.
  • 3
    Holmes MV, Hunt BJ, Shearer MJ. The role of dietary vitamin K in the management of oral vitamin K antagonists. Blood Rev 2012; 26: 114.
  • 4
    Laliberte R, Chakrabarti S, Brodeur J. The influence of fasting on the disposition of warfarin in rats. J Pharmacol Exp Ther 1977; 200: 4451.
  • 5
    Laliberté R, Chakrabarti S, Brodeur J. The influence of fasting and stress on the response of rats to warfarin. J Pharmacol Exp Ther 1976; 196: 194203.
  • 6
    Schulman S, Kearon C, Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost 2005; 3: 6924.
  • 7
    Kong MC, Lim TG, Ng HJ, Chan YH, Lee LH. Feasibility, cost-effectiveness and patients' acceptance of point-of-care INR testing in a hospital-based anticoagulation clinic. Ann Hematol 2008; 87: 90510.
  • 8
    Sobieraj-Teague M, Daniel D, Farrelly B, Coghlan D, Gallus A. Accuracy and clinical usefulness of the Coaguchek S and XS Point of Care devices when starting warfarin in a hospital outreach setting. Thromb Res 2009; 123: 90913.
  • 9
    Lawrie AS, Hills J, Longair I, Green L, Gardiner C, Machin SJ, Cohen H. The clinical significance of differences between point-of-care and laboratory INR methods in over-anticoagulated patients. Thromb Res 2012; 130: 1104.
  • 10
    Rosendaal FR, Cannegieter SC, van der Meer FJ, Briët E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993; 69: 2369.
  • 11
    Ramadan: health effects of fasting. Lancet 2009; 374: 588.
  • 12
    Ziaee V, Razaei M, Ahmadinejad Z, Shaikh H, Yousefi R, Yarmohammadi L, Bozorgi F, Behjati MJ. The changes of metabolic profile and weight during Ramadan fasting. Singapore Med J 2006; 47: 40914.
  • 13
    Hawk TL, Havrda DE. Effect of stress on international normalized ratio during warfarin therapy. Ann Pharmacother 2002; 36: 61720.
  • 14
    Reynolds MW, Fahrbach K, Hauch O, Wygant G, Estok R, Cella C, Nalysnyk L. Warfarin anticoagulation and outcomes in patients with atrial fibrillation: a systematic review and meta-analysis. Chest 2004; 126: 193845.
  • 15
    Wan Y, Heneghan C, Perera R, Roberts N, Hollowell J, Glasziou P, Bankhead C, Xu Y. Anticoagulation control and prediction of adverse events in patients with atrial fibrillation: a systematic review. Circ Cardiovasc Qual Outcomes 2008; 1: 8491.