Introduction: Obesity has emerged as a significant public health problem in all ages in the United States. The age-adjusted prevalence of obesity and severe obesity were 42.2% and 9.2% respectively among adults aged 20 or older from 2017-2018. Overall, non-Hispanic black and Hispanic adults and youth had a higher prevalence of obesity compared with other race and Hispanic-origin groups. Obesity prevalence was lower among non-Hispanic Asian men and women compared with other race and Hispanic-origin groups based on National Center for Health Sciences- Center for Disease Control. Factor VIII (FVIII) and von Willebrand factor (vWF) are both coagulation factors that are acute phase reactants. Obesity has been associated with metabolic syndrome and resultant chronic inflammation. Obesity has previously been linked to increasing FVIII levels, however, studies have not been able to confirm a linkage between VWF and body mass index (BMI). Methods: This is a secondary analysis of a single-site retrospective review of the electronic health record of patients that had von Willebrand profiles [VWF:Antigen (Ag), VWF:Activity (Act), and/or FVIII activity] obtained within 24 hours of a hemoglobin level who had no bleeding disorder diagnosis. BMI was recorded when available in the electronic medical record at the time of laboratory testing and correlation between BMI and FVIII/vWF levels was assessed. Data was retrieved from January 1, 2012, through June 1, 2021. Results: We identified 4552 individuals who had VWF profile testing that had an associated hemoglobin level in a 24-hour period and did not carry a diagnosis of VWD. Among all subjects, a total of 2152 subjects were found to have BMI measurements, 1294 subjects <85th percentile, 379 subjects ≥85th - 95th percentile and 479 subjects ≥ 95th percentile. Demographics for the cohort with BMI measurements were White (N=1,673; mean BMI percentile 67.4±29.4), Black (N=377; mean BMI percentile 72±27.3), Asian (N=46, mean BMI percentile 48.0±33.9). There was a statistically significant difference in BMI by race (P<0.001). VWF antigen (P<0.005) and FVIII level (P<0.005) were significantly and positively correlated with increased BMI levels. Similarly, VWF activity/antigen ratio was negatively correlated (P<0.005) with increasing BMI. Further statistical modelling was performed, controlling for hemoglobin, BMI percentile, race, and age. BMI percentile remained a determinant for VWF antigen and FVIII levels. VWF antigen levels were statistically different in those with BMI <85th and ≥ 95th percentile (P<0.0001) and ≤85th-95th and ≥ 95th percentile (P<0.01). FVIII levels were significant between BMI ≤85th and ≥ 95th percentile (P<0.0001). We also examined the cohort for potential differences in VWF and FVIII levels based on race. Black race was found to be associated with higher VWF antigen levels among all races and found to be statistically significant when compared to White race (P<0.0001). Black race was also associated with a higher FVIII level versus White race (P<0.0001). Further statistical modeling was also done for race, controlling for BMI, hemoglobin, and age. VWF antigen levels were statistically significant among races (P<0.0001) with Black race having the highest mean level. Similarly, Race was a significant factor for VWF activity/antigen ratio (P<0.0001) with White race having the highest and Black race the lowest mean ratio. Race was also statistically significant (P<0.0001) for FVIII levels, with Black race having the highest levels and lowest levels in the all-other race category. Conclusions: BMI and race were independently found to be determinants of VWF:Ag and FVIII levels. As patients presenting for a bleeding disorder evaluation may be overweight or obese, the confounding effect of increased BMI should be considered during the diagnostic evaluation. We also observed racial differences in VWF antigen and FVIII levels. As this was antigen testing, it should not be affected by the known racial differences in sensitivity to ristocetin cofactor testing. Further studies to confirm these findings and the potential confounding effect of BMI are needed.
Read full abstract