Is Age Related Fibrinogen Affected by Warfarin?.

Introduction: There is published evidence which indicates that advancing age may be associated with higher plasma concentrations of fibrinogen. There is also evidence that derived fibrinogen values are significantly higher than Clauss measurements and that this discrepancy is greater in patients receiving warfarin. The purpose of this study was to determine whether age related derived fibrinogen levels are similar in both warfarin and non-warfarin groups.

Methods: Venous samples were collected into siliconised glass B-D Vacutainers containing tri-sodium citrate (Ref: 367691) from 1000 patients receiving long term warfarin treatment and an equal number of age-matched patients not receiving warfarin. Genders were equally represented in both groups. Patients in both groups were categorized into 5 years age bands as follows: <40 n=23: 40–44 n=20: 45–49 n=43; 50–54 n=74: 55–59 n=113: 60–64 n=155: 65–69 n=178: 70–74 n=191: 75–79 n=124; 80–84 n=56: 85–89 n=23. Derived fibrinogen was measured in each patient on an ACL300R coagulometer (I L) within 1 hour of collection using IL PT-FIB HS Plus reagent and following the manufacturer’s protocol. Appropriate CLSI guidelines were followed throughout. A normal probability plot of the data was performed to confirm that it did not deviate too much from the normal distribution.

Results: The T-test for independent samples using the separate variance estimate showed that there was a statistically significant difference in the mean fibrinogen between patients on warfarin and those not on warfarin (p<0.05) in each group except for the last (85–89 years). There was a statistically significant difference (ANOVA) in the fibrinogen levels of patients of different age in both warfarinised and non-warfarinised groups (p<0.05). The modified least significance procedure in the ANOVA test showed that in the non-warfarin group, most of significant difference in fibrinogen between the different age groups is contributed by the difference between patients under 50 years of age. In the non-warfarin group, it requires an age gap of at least 20 years for the difference in fibrinogen to be statistically significant but in the warfarin group, it only requires an age gap of ten years (p<0.05). Both Linear Regression and Cross Tabulation indicate that the relationship between fibrinogen and age does not vary whether or not the patient is on warfarin. These also show that the effect of age on fibrinogen is not affected by warfarin treatment.

Conclusion: Differences or correlations detected in this analysis are of statistical significance but not necessarily clinically significant. Placing age and warfarin treatment in the same model shows that variations in fibrinogen have to be explained by other factors (e.g. technical) not included in the study as only 12% of the error in predicting fibrinogen levels can be reduced by knowing both the age and status of warfarin treatment in individual patients.