Red-haired individuals are known for their reputation of increased bleeding during surgical procedures and in women after childbirth . However, only a few sound scientific studies have been performed to assess the association between red hair and bleeding. In a small study performed by Liem et al.,  women with red hair indeed reported a slightly increased rate of bruising. Reid et al. have investigated a limited number of coagulation parameters in red-haired vs. black-haired males. They found all laboratory results to be within the normal range . Another previous study examined both primary and secondary hemostasis tests, including activated partial thromboplastin time (APTT), prothrombin time (PT), platelet aggregometry and platelet function analysis (PFA-100), in 25 healthy women with red hair and compared this with a control group of 26 dark-haired women . Also in this study no differences were observed between the two groups for these global tests. However, it is known that the PFA-100 is not very sensitive to detect mildly decreased von Willebrand factor (VWF) levels. In addition, they observed a slightly lower high-dose, ristocetin-induced platelet aggregation (RIPA) in red-haired (87%) vs. in black-haired women (93%, P = 0.06). This observation led Favaloro  to propose that VWF may play a role in the red-hair associated bleeding phenotype, as the RIPA test is dependent on plasma VWF. In both the above-mentioned coagulation studies, VWF levels were not measured [2,3], whereas it is known that reduced levels of VWF may confer a risk factor for bleeding, even in individuals that have not been diagnosed with von Willebrand disease . This led us to assess the association between hair color and VWF levels in a large cohort of nearly 4000 healthy individuals.
For this study, we included participants from the Rotterdam Study (RS), an ongoing, prospective, population-based cohort study among individuals of 55 years and older living in a suburb in the city of Rotterdam, the Netherlands, of whom > 90% are Caucasian . Out of 3957 individuals, plasma was available for VWF determination and hair color (when young) was known for 97%.
Fasting venous blood samples were collected in citrated tubes and plasma was stored at −80 °C. VWF antigen (VWF:Ag) was determined with an in-house ELISA with polyclonal rabbit anti-human VWF antibodies and horseradish-peroxidase-conjugated anti-human VWF antibodies (DakoCytomation, Glostrup, Denmark) for catching and tagging, respectively. The intra-assay coefficient of variation was 5.8% and the inter-assay coefficient of variation was 7.8% . Hair color when young was self-reported and classified as red, fair/blond, black or brown.
The characteristics of the study population are shown in the Table 1. The mean (standard deviation [SD]) age of the total study group was 65.7 (7.0). Of the included individuals 221 (57.8%) were female.
|Red hair||Fair hair||Brown hair||Black hair|
|Number (%)||104 (2.6%)||860 (21.7%)||2511 (63.5%)||369 (9.3%)|
|Age (years)||65.2 (6.7)||66.5 (7.1)||65.6 (6.9)||65.2 (6.9)|
|Female gender (%)||55.8%||57.3%||59.2%||49.9%|
|Body mass index (kg m−2)||26.4 (3.7)||26.3 (3.5)||26.3 (3.5)||26.1 (3.4)|
|VWF:Ag in IU dL−1*||1.39 (0.40–3.30)||1.41 (0.30–8.40)||1.36 (0.30–7.80)||1.36 (0.20–5.80)|
VWF:Ag levels for the different hair colors are given in the Table 1. No clear difference was observed between the groups. Red-haired individuals (n = 104) had VWF:Ag mean (range) levels of 1.39 IU dL−1 (0.4–3.3). VWF:Ag was 1.41 IU dL−1 (0.3–8.4) for fair hair (n = 860), 1.36 IU dL−1 (0.3–7.8) for brown hair (n = 2511) and 1.36 IU dL−1 (0.20–5.80) for black-haired individuals (n = 369). Using a linear regression model, VWF levels were significantly dependent upon age (beta = 0.02, SE 0.001, P < 0.001) and sex (beta = −0.06, SE 0.02, P = 0.001), but not of hair color (beta = 0.03, SE 0.06, P = 0.65). When blood group was included in the linear regression model the beta increased to 0.06, however, this was still not significant (P = 0.323). Low VWF levels, defined as a level < 5% of all individuals, were evenly distributed over all hair colors (P = 0.49, corrected for blood group). We observed a slightly higher frequency of blood group O in the red-haired individuals compared with the individuals with other hair colors; however, this was not statistically significant (P = 0.07).
Recently, a genome-wide association (GWA) study identified two genetic loci close to ASIP and MC1R that are associated with red hair color . We examined whether the single nucleotide polymorphisms (SNPs) in these two loci that were significantly associated with red hair in the GWA study were also associated with red hair color and VWF levels in our population. We found that at least one SNP in the ASIP locus (P-value < 1.3 × 10−4) and 10 SNPs in the MC1R locus (P-value < 3.9 × 10−6) were significantly associated with red hair color. These SNPs explain 24% of the variation in hair color in our population. When we examined the association of these SNPs with VWF levels, the smallest P-value found was 0.11 for the ASIP locus and 0.07 for the MC1R locus. This indicates that genetic variants associated with red hair color do not contribute to the variation in VWF antigen levels. This suggests that red hair color and VWF protein levels are not genetically linked.
Studies performed so far on the association between red hair and more pronounced bleeding have not yet revealed that this is caused by changes in coagulation parameters [2,3]. Also our current study does not support such an association. However, other factors outside the coagulation mechanism may influence the bleeding risk in individuals undergoing surgery, including vascular factors. Several studies have reported an increased risk of hernia formation in red-haired individuals, which may be linked to collagen synthesis [1,9]. Others have reported on the increased requirements for anesthesia of red-haired individuals . Both these observations suggest that other underlying mechanisms for the presumed association between bleeding and hair color may exist.
It has previously been suggested that the keratin structure may have a possible role in bleeding in red-haired individuals . Because of the analogy of poly-protein subunits linked by Cys–Cys bonds both in keratin and VWF, the control of the keratin structure may also be linked to the control of the plasma VWF structure and function. In addition, the keratin gene is located on chromosome 12, (12q12-q13) as is the VWF gene (12p13.3). Therefore, using our VWF database, we additionally made a regional plot of this region and observed a signal (rs662843) with a P-value of 7 × 10−4, which was, however, given the number of studied SNPs below the significance threshold of 3.5 × 10−5. Therefore, no genetic association between VWF and keratin was observed.
A limitation of the present study is that we only studied VWF:Ag levels, and it is still possible that VWF activity (collagen-binding and/or ristocetin co-factor activity) may be lower in red-haired individuals. Based on findings that collagen defects may be present in these individuals the collagen-VWF interaction may be reduced .
We conclude that VWF antigen levels and genetic variants that determine VWF antigen levels do not explain the presumed increased bleeding phenotype of red-haired individuals.