PETER J. D’ADAMO and GREGORY S. KELLY
Copyright 2000-2009 All Rights Reserved. Unauthorized reproduction prohibited by law.
This monograph will attempt to organize and explore these associations in some detail.
Functional and Genetic Factors Involved in ABH Secretion
The term “ABH secretor,” as used in blood banking, refers to secretion of ABO blood group antigens in fluids such as saliva, sweat, tears, semen, and serum. If people are ABH secretors, they will secrete antigens according to their blood groups. For example, group O people will secrete H antigen, group A people will secrete A and H antigens, etc. Soluble (secreted) antigens are called substances. To test for secretor status, an inhibition or neutralization test is done using saliva. The principle of the test is that if ABH antigens are present in a soluble form in a fluid (e.g., saliva) they will neutralize their corresponding antibodies and the antibodies will no longer be able to agglutinate red cells possessing the same antigens.
One of the primary differences in physiology between secretors and non-secretors has to do with qualitative and quantitative differences in components of their saliva, mucus, and other body secretions. ABH secretion is controlled by two alleles, Se and se. Se is dominant and se is recessive (or amorphic). Approximately 80% of people are secretors (SeSe orSese).
In the most rudimentary sense, the secretor gene (FUT2 at 19q13.3) codes for the activity of the glycosyltransferases needed to assemble aspects of both the ABO and Lewis blood groups. This it does in concert with the gene for group O, or H (FUT1). These enzymes are then active in places like goblet and mucous gland cells, resulting in the presence of the corresponding antigens in body fluids.(1)
The H antigens are indirect gene products expressed as fucose-containing glycan units, residing on glycoproteins or glycolipids of erythrocyte membranes or on mucin glycoproteins in secretions and are the fucosylated glycans substrates for glycosyltransferases that give rise to the epitopes for the A , B and Lewis blood group antigens. The major difference between the two genes is in their pattern of expression: the FUT1 (H) gene is expressed predominantly in erythroid tissues giving rise to FUT1 (H enzyme) whose products reside on erythocytes, whereas the FUT2 (Secretor) gene is expressed predominantly in secretory tissues giving rise to FUT2 (Secretor enzyme) and to products that reside on mucins in secretions.
When alleles of both genes fail to express active enzymes, individuals bearing them, in homozygous state, lack the substrates for the A or B glycosyltransferases and do not express the A and B epitopes.
Relationship of ABH Secretor Status and Lewis System
Since FUT1 provides the glycans necessary for glycosyltransferases conversion into the Lewis antigen in addition to ABH, the Lewis blood group determinants are structurally related to determinants of the ABO and the H/h blood group systems and the outcome of Lewis typing can also often be used for the de facto determination of ABH secretor status. In the presence of FUT2 alleles that express type 1 H determinants, the phenotype will be Le (a-b+) but individuals in whom the FUT2 gene is not expressed will be (Le a+b-).
ABH secretors are almost always Lewis (a-b+) since they convert all their Lewis (a) antigen into Lewis (b). ABH non-secretors are always (Lewis a+b-) since they lack the FUT2 dependent glycosyltransferase to accomplish this. A small section (1-4% of the population dependent on race) will be Lewis Double Negative (LDN; Lewis (a-b-)) and for which Lewis typing cannot be used to determine ABH secretor status. In these individuals determination via saliva is necessary. However, it may be helpful to think of LDN individuals as a special category of non-secretor, since they do lack the Lewis b antigen (like the traditional ABH non-secretors). In most instances LDNs share the same metabolic consequences as ABH non-secretors, and in a few, such as cardiovascular disease and insulin resistance, actually have the most severe variations.
|Lewis Phenotype||ABH Secretor Status|
Le (a+b-) which has Lewis a antigen but not Lewis b
Always ABH non-secretor
Le (a-b+) which has Lewis b antigen but not Lewis a
Always ABH secretor
Le (a-b-) having neither Lewis a nor Lewis b
Lewis outcome not a determinant of ABH secretor status. However, this variant is associated with its own unique metabolic consequences.
Table 1: Lewis blood types and their relationship to ABH secretors/non-secretor status.
Although ABH secretor status is often thought of as an all or none situation, this is generally not the case. In some ABH non-secretors (known as partial or weak secretors) there will often be some form of active A or B blood group substance in the saliva; however, the quantity and quality of these substances is greatly reduced, predisposing them to similar functional problems as other non-secretors. (3,4)
Antigenic Structures in Fluid Secretions
There are several advantages in having large quantities of blood type antigens (both ABO and Lewis) secreted into saliva. First, salivary carbohydrate structures found in mucins can aggregate some oral bacteria and also constituents of pellicle and plaque. Since saliva of secretors contains substantially more diversity and total carbohydrate than non-secretor mucins, this places secretors at a bit of an advantage. Second, these same blood type carbohydrate structures, because of the noted sweet tooth of many dietary lectins, might actually place secretors at a significant advantage with respect to binding some blood type specific dietary lectins.
In the gastric mucosa of healthy individuals, the normal mucosa of secretors is characterized by a uniform distribution of blood type antigens in the pits. Healthy mucosa of non-secretors shows little staining for these blood type antigens, but, instead, demonstrates significant quantities of the I(Ma) antigen. This tendency to express the I (Ma) antigen will subsequently have an impact on antibody capabilities, as we will see when we discuss immunity. (5)
Brush Border Hydrolases
ABO blood group determines much of the enzyme activity in the tissue (brush-border) of the intestine. At least six intestinal hydrolases have ABO blood group antigenic determinants that are directly related to ABO blood group. Basically, the intestinal glycoproteins of blood group A and B individuals express A or B antigens, while blood group O subjects express the H determinant. The expression of these ABH antigens is under the control of the secretor gene; so these ABH antigens are not detected in the hydrolases of non-secretor subjects. (6) ABH secretors have greater quantities of free ABH antigens in the makeup of their intestinal secretions; this has significant effects on bacterial and lectin adherence to the gut microvilli.
Intestinal Alkaline Phosphatase Activity
The activity of intestinal alkaline phosphatase and serum alkaline phosphatase is strongly correlated with ABH secretor phenotypes. Independent of ABO blood group, ABH non-secretors have lower alkaline phosphatase activity than ABH secretors. It has been estimated that the serum alkaline phosphatase activity of non-secretors is only about 20% of the activity in the secretor groups. (7-10)
The intestinal component of alkaline phosphatase is involved with both the breakdown of dietary cholesterol and the absorption of calcium. The differences in intestinal alkaline phosphatase are almost exclusively related to one fraction of the intestinal alkaline phosphatase. Normal molecular mass intestinal alkaline phosphatase (NIAP) is present in the serum of both secretors and non-secretors regardless of ABO blood group. However, the high molecular mass intestinal alkaline phosphatase only appears in serum of Lewis (a-b+) blood group secretors. (11)
It should be mentioned that in addition to ABH secretor status, ABO polymorphism is also linked to the levels and persistence of intestinal alkaline phosphatase. (88) Numerous studies have associated group O individuals with the highest alkaline phosphatase activity and group A the least. (89)
These findings suggest that the link between group O individuals and adaptation to cholesterol-containing foods in the diet (such as meats) reaches its greatest accommodation in group O secretors. Conversely, group A non-secretors would have the lowest levels of intestinal alkaline phosphatase and the greatest difficulties in handling dietary fat. In addition, one study has implied that the group A antigen itself may inactivate IAP. (90)
The role of ABO blood group in determining some of the bacteria making up a healthy GI ecosystem is particularly strong in ABH secretors. Since ABH secretor status and ABO blood group dictate the presence and specificity of A, B, and H blood group antigens in human gut mucin glycoproteins, this can influence the populations of bacteria capable of taking up local residence. This occurs because some of the bacteria in the digestive tract are actually capable of producing enzymes that allow them to degrade the terminal sugar of the ABH blood type antigens for a constant food supply.
For example, bacteria capable of degrading blood group B antigen produce enzymes that that allow them to detach the terminal alpha-D-galactose and use this sugar for food. Blood group A degrading bacteria would have similar capabilities with respect to N-acetylgalactosamine. Group B secretors produce greater levels of B-degrading than A- or H-degrading activity, and A secretors produce greater levels of A-degrading than B- or H-degrading activity. Because of this capability, the bacteria that use ABH antigens for food have a competitive advantage and can thrive in the environment created by the preconditioning of ABH secretions.
Although comparatively small populations of bacteria produce blood group-degrading enzymes (estimated populations are 10(8) per g); the quantity of these bacteria are several orders of magnitude greater in different blood types and are much more stable residents. For example, B-degrading bacteria have a population density of about 50,000-fold greater in blood group B secretors than in other subjects. Similar bacterial specificity and enzyme activity is found among other blood types. (12,13)
Breast Milk Components
Significant variations in the carbohydrate residues in human breast milk are found depending on the mothers ABO, Lewis, and Secretor blood types. During the first week of lactation the ability to produce neuraminyloligosaccharides is linked to the ABH secretor groups. And the ability to produce oligosaccharides with Le (a) or Le (b) characteristics is linked to Lewis and Secretor systems. The consequences of this are that secretors will produce higher levels of N-acetylneuraminic acid and lower levels of galactose in their breast milk than non-secretors. In the ABH secretor groups, blood type A and O secretors also have higher N-acetylglucosamine contents than B and AB secretors (p less than 0.001), while the A and B secretors have higher galactose levels. The Lewis secretor groups are also distinguished by a significantly higher level of fucose. The ABH (+), Le (a-b-) group had higher lactose contents than the other groups. (28)
Maternal ABO Blood Group, ABH secretor status and Lewis Phenotype
|Composition of Breast Milk|
A or O, Secretors,
High amounts of N-acetylneuraminic acid and N-acetylglucosamine
Lower levels of galactose
Presence of Le(b) and either A or O substances
B or AB, Secretors,
High amounts of N-acetylneuraminic acid
Presence of Le(b) and either B or AB substances
Lowest amounts of N-acetylneuraminic acid
Presence of Le(a) & absence of ABO substances
ABO, Lewis negative,
No Lewis substances
Quantities of ABO substances, galactose, N-acetylneuraminic acid, & N-acetylglucosamine can not be estimated
Table 2. Differences in carbohydrate composition by ABO and Lewis blood groups and ABH secretor status.
ABO blood group impacts the clotting ability to a significant degree, in fact, it has been estimated that a significant fraction (30%) of the genetically determined variance in plasma concentration of the von Willebrand factor antigen (vWf) is directly related to ABH determinants. As a rule, it is blood group O individuals who have the lowest amount of this clotting factor. (29)
ABH non-secretors are reported to have shorter bleeding times and a tendency towards higher factor VIII and vWf. This relationship appears to be another example of blood type synergy between ABO and Secretor/Non-secretor phenotypes. In fact, secretor genetics appear to interact with ABO genetics to influence as much as 60% of the variance of the plasma concentration of vWf, with secretors (Le (a- b+)) having the lowest vWf concentrations. (30,31)
Among persons belonging to blood group O (the blood type most likely to have problems with clotting), the lowest concentration of vWf:Ag and VIII:Ag is found in the group O secretors. While blood group O non-secretors will have a higher concentration of both vWf:Ag and factor VIII antigen (VIII:Ag), providing them with a better capability for clotting. (32)
Among blood groups A, B, and AB, also having the Lewis (a- b-) phenotype is associated with the highest degree of clotting factors. In white men with these blood types, the Lewis (a-b-) phenotype, infers significantly higher levels of factor VIII and von Willebrand factor. Among black men with blood type A, B, or AB, and phenotype Lewis (a-b-), a similar trend is found with these individuals having the highest values for factor VII and von Willebrand factor. In women with blood type A, B, or AB, and phenotype Le(a-b-) a correlation exists for higher levels of factor VIII.
Le (a- b-) highest activity of factor VIII and vWf
Shortest bleeding times (especially in A, B and AB)
Le (a+ b-) intermediate activity
Shorter bleeding times (especially for O)
Le (a- b+) lowest activity of factor VIII and vWf
Longest bleeding times (especially for O)
Table 3: Lewis Blood Type and Clotting Factors
Based upon this research, researchers have suggested that the Le (a-b-) phenotype (and blood groups A, B, and AB especially), by virtue of their association with raised levels of factor VIII and von Willebrand factor, might be at a higher risk for future thrombotic and heart disease. (33)
In all blood groups, the average amount of cavities is lower for ABH secretors than for non-secretors. This difference is most significant for smooth surface areas of the teeth. Also, secretors of blood group A had the lowest numbers of cavities. (34)
DIABETES, HEART DISEASE, & METABOLIC SYNDROME X
ABH non secretors, and especially Lewis negative individuals, are at a greater risk of developing diabetes (especially adult onset diabetes); and they might be at a greater risk of developing complications from diabetes. Findings suggest that an increased proportion of non-secretors are found among patients with diabetes, particularly of the insulin-dependent diabetes type. (14,15)
The Lewis negative (Le a-b-) red blood cell phenotype appears to confer the greatest risk of developing diabetes. This blood type is observed more than three times more frequently (29%) in diabetics irrespective of their clinical type. Non-diabetics categorized as low insulin responders to glucose are also significantly more likely to be Lewis negative. (16)
Among individuals with juvenile diabetes mellitus, the prevalence of severe retinopathy (a side effect of diabetes) is lower in ABH secretors than in the ABH non-secretor group. (17)
Data allows the conclusion that the ABH non-secretor phenotypes are a risk factor for myocardial infarction, this is particularly true for recessive Lewis blood types and even more so among men than women. ABH secretors seem to have been given a bit of genetic resistance against heart disease while Lewis negative individuals appear to be at the highest risk for CHD.
This finding was reported in the Copenhagen Male study and replicated in the NHLBI Family Heart Study. Eight percent of men with the Lewis (a-b-) phenotype had a history of non-fatal myocardial infarction (among Lewis positive men the frequency was only 4%). But even worse, research showed that men with Lewis (a-b-) had an increased risk of death from ischemic heart disease (IHD) (IHD case fatality rate (RR = 2.8 (1.5-5.2), P = 0.01)) compared with others. Adjusted for age, relative risk climbed even higher to 4.4 (1.9-10.3), P < 0.001, and for all causes of mortality RR = 1.6 (1.0-2.6), P < 0.05. (18)
Results from the NHLBI Family Heart Study also showed a higher risk of coronary heart disease (odds ratio was 2.0 (95% confidence interval = 1.2 to 3.1) for Lewis (a-b-) versus other Lewis groups. Triglycerides were significantly higher in the Lewis (a-b-) subjects. Among women, there was also a trend towards increased risk of CHD among Lewis negative phenotypes; however, the trend is dramatically weaker than among male subjects. (19)
Additional research has also duplicated these results, supporting and adding to the weight of evidence linking Lewis negative phenotype Lewis (a-b-) as a marker of high risk for the development of ischemic heart disease. Even excluding the Lewis negative phenotype, the secretor phenotype Lewis (a-b+) was found to be a genetic marker of resistance against the development of ischemic heart disease, while ABH non-secretor status is a risk factor predisposing individuals towards heart disease. (20)
Protective Effects of Alcohol
In men Lewis (a-b-), the Lewis negative or excessive phenotype, a group genetically at high risk of ischemic heart disease (IHD), alcohol consumption seems to be especially protective. In the Copenhagen study, researchers found that drinking alcohol was the only risk factor that had an interaction with Lewis negative blood type and that alcohol could strongly modify risk in an inverse (so hence positive) manner. There was a significant inverse dose-effect relationship between alcohol consumption and decreasing risk. (21)
Alcoholism and Alcohol’s Protective Benefits
Paradoxical with the cardiovascular benefits of alcohol in Lewis negative individuals, several large studies have associated alcoholism with ABH non-secretor status. (22,23)
Metabolic Syndrome X
Data suggest that Lewis (a-b-) men exhibit features of the insulin resistance syndrome or syndrome X, including a tendency to prothrombic metabolism and higher levels of BMI, SBP, triglycerides, and fasting levels of serum insulin and plasma glucose. These same relationships are not as strong for women.
A group of metabolic problems comprised of insulin resistance, elevated plasma, lipid regulation problems (elevated triglycerides, increased small low-density lipoproteins, and decreased high-density lipoproteins), high blood pressure, a prothrombic state, and obesity (especially central obesity or a predisposition to gaining weight in the abdomen) combine to form “Metabolic Syndrome X” (MSX). This cluster of metabolic disorders seems to promote the development of diabetes (adult onset type II), arteriosclerosis, and cardiovascular disease. And while insulin resistance might lie at the heart of the problem, all of these metabolic disorders appear to contribute to health problems.
Because of the associations with non-secretor status and both diabetes and heart disease, many different researchers have explored the connection between a metabolic syndrome called “Syndrome X” and Lewis and non-secretor blood types. Just as is the case with diabetes and heart disease, individuals with Lewis (a-b-) phenotype are most predisposed to MSX. It has even been hypothesized that Lewis (a-b-) men and syndrome X share a close genetic relationship on chromosome 19 and that the Lewis (a-b-) phenotype is a genetic marker of the insulin resistance syndrome. (24)
As we will discuss in the next section on clotting, non-secretors and especially Lewis negative individuals, are especially prone to prothrombic metabolism (a tendency to form clots more readily and to have slower bleeding times). The tendency to higher triglycerides was mentioned when we discussed heart disease. (25)
Researchers have also investigated Lewis blood types as part of the Copenhagen Study, and they found very supportive evidence of trends toward metabolic differences. Compared to all other men, the Le (a-b-) men had a significantly higher systolic blood pressure (6 mm Hg, P = .0024). They also had higher values of body mass index (8%, P = .016), total body fat mass (25%, P = .015), fasting values of serum insulin (32%, P = .006), serum C-peptide (20%, P = .029), and plasma glucose (8%, P = .003). These trends, while consistent for men, were not as strong for women. (26)
Evidence suggests that ABH non-secretors have lower levels of IgG. (35,36) In tests of 202 Caucasians researchers found IgA concentrations to be significantly lower in non-secretors than in secretors. (37,38) This seems to imply that the ABH non-secretor state is associated with a “Defense In Depth” strategy (i.e. let the invader in and attempt to destroy it internally) versus the ABH secretor state, which implies a “Preclusive Strategy” (i.e. wall out the invader and don’t allow entrance in the first place.) For example, the free ABH antigen on the mucosa barriers of ABH secretors acts as an effect anti-adhesive mechanism against ABH specific bacterial fimbrae lectins.
On the other hand, the ability to secrete relatively different concentration of the components of the blood group substances as determined by secretors/non-secretor genetics seems to affect phagocytic activity of the leucocytes in a manner that actually places non-secretors at somewhat of an advantage. In general, leukocytes of non-secretors have substantially greater ingestion power as compared to secretors. Although this ability appears to be across the board for all non-secretors, blood group O and B non-secretors have the greatest advantage and highest phagocytic activity. (41) Perhaps this is a compensatory mechanism for their more limited antigenic barrier in their body fluids and secretions.
Results suggest that the level of anti-I in the serum of normal individuals may be affected by the donor’s ABO group, secretor status and sex. For individuals with blood group O, B and AB secretors have higher levels of an antibody presumed to be auto-anti-I (cold hemagglutinin). The level of this antibody is usually even higher among non-A female secretors than for males. (39)
Researchers have found that in individuals with insulin dependent diabetes mellitus, the mean level of C3c for non-secretors is significantly lower than that found for secretors. The level of C4 among ABH non-secretors was also significantly lower than that of ABH secretors. (40)
The genetics of the ABH secretor/non-secretor system interact to alter an individual’s risk for ulcers. In several studies, non-secretors of ABH substances have been found to have a significantly higher rate of duodenal and peptic ulcers. (42,43)
In fact the Copenhagen study found that the lifetime prevalence of peptic ulcer in men who were ABH non-secretors was 15% (statistically 15% of ABH non-secretors will have an ulcer at some point in their lives). And, the attributable risk of peptic ulcer in men who were Lewis (a + b-) or ABH non-secretors, with blood group O or A phenotypes was 37%. (44)
Overall, the relative risk of gastroduodenal disease for non-secretors compared with secretors is 1.9 (95% confidence interval). Duodenal ulcer patients are more likely to be non-secretors, and being a non-secretor acts as a multiplicative risk factor with the gene for hyperpepsinogenemia I to impact the risk of duodenal ulcer. (45,46)
Because of the increased prevalence of ulcers among non-secretors researchers have suggested that secretor status might influence bacterial colonization density or the ability of H. pylori to attach to gastroduodenal cells. With regards to the overall interaction with H. pylori infection, non-secretor status is generally considered to be a separate independent risk factor for gastroduodenal disease in addition to H. pylori infection; however, there is more to this story, and, in fact some interesting interactions between secretor status, Lewis genetics, and H. pylori. (47)
Because non-secretors are limited in their ability to secrete the Lewis (b) blood group antigen into the mucus secretions of their digestive tract, it has been proposed that they be at a competitive disadvantage from preventing H. pylori attachment. In fact, the Lewis (b) antigens have been found to act as somewhat of a preferential target for H. pylori attachment. Thus, lack of Lewis (b) in mucosal fluids of ABH non secretors might indirectly contribute to colonization by H. pylori. (48-50)
In a simplified sense, when the Lewis (b) antigen is free floating in the mucus, it probably acts to bind up some of the H. pylori before it can contact and attach to host tissue. In essence, being an ABH secretor provides an ability to put some biological decoys or metabolic chaff out into the gastric secretions that is very specific for H. pylori. Also, in ABH non-secretors the immune response against H. pylori appears to be lower and H. pylori appears to attach with higher aggressiveness and cause more inflammation. (51)
Individuals with Lewis (a+b-) ABH non-secretor phenotype also show a significantly higher proportion of the H. pylori-seronegative subjects and a lower IgG (H. pylori immunoglobulin G (IgG) antibody) immune response to H. pylori antigens as compared with the individuals of Lewis (a-b+)/secretor phenotype.
Evidence also indicates that 100% of non-secretors with duodenal ulcers culture positive for H. pylori infection. However, among non-secretors with gastric ulcer, H. pylori is found in only about 12.5% of the cases. This is not observed among secretors, who are nearly equally likely to have H. pylori infection in either gastric or duodenal ulcer. (52)
Bacteria Urinary Tract Infections
ABH non-secretors are at a greater risk for recurrent urinary tract infections (UTI) and are much more likely to develop renal scars. This susceptibility is even greater among the Lewis negative subset (Le (a-b-)). The ABH secretor phenotype conveys a measure of protection; cutting the risk of recurrent UTI by greater than 50% and dramatically decreasing the likelihood that renal scars will develop.
ABH non-secretors appear to be at extra risk for recurrent urinary tract infections. In one study of women with recurrent UTI, 29 % of the women were the Lewis (a+ b-) non-secretor phenotype, while another 26% of the women were Lewis (a- b-) recessive phenotype. When the women with ABH non-secretor and recessive phenotypes were combined and considered collectively, the odds ratio (an estimate of relative risk of recurrent urinary tract infection) for those without the secretor phenotype (Lewis (a-b+) was 3. (53-57)
A form of synergy also appears to exist between UTI risk, secretor status and the lack of ability to create anti-B isohemagglutinin. Essentially, blood group B and AB and the non-secretor phenotype seem to work together to increase the relative risk of recurrent UTI among these women. (58) Evidence also indicates that women and children with renal scarring subsequent to recurrent UTI and pyelonephritis are more likely to be ABH non-secretors. (59-61) As many as 55-60% of all ABH non-secretors have been found to develop renal scars, even with the regular use of antibiotic treatment for UTI whereas as few as 16% of ABH secretors will develop similar renal scarring. (62)
This tendency to scarring does not seem to be dictated as much by the aggressiveness of the bacterial infection, but by the more aggressive inflammatory response created by ABH non-secretors against the bacterial infection. The levels of C-reactive protein, erythrocyte sedimentation rate and body temperature are significantly higher in non-secretors than in secretors (p less than 0.04) with recurrent UTI. As a consequence, non-secretors seem to self inflict to a degree the renal scarring secondary to their acute phase inflammatory response. (63)
The genetically determined inability to secrete the water-soluble glycoprotein form of the ABO blood group antigens into saliva and other body fluids is a recognized risk factor for Neisseria meningococcal disease. ABH non-secretors are consistently over represented among individuals contracting this infection. This over representation is even greater among individuals who are carriers of the infection. (64) Secretory immune capabilities and other factors appear to contribute to the relative protection against colonization by meningococci enjoyed by ABH secretors. ABH non-secretors typically have lower levels of anti-meningococcal salivary IgM, and if to add insult to injury, both the IgA and IgM antibodies produced by ABH secretors are more effective at providing protection against this microorganism. (65)
ABH non-secretors are much more likely to be carriers of Candida sp. and to have problems with persistent Candida infections. Blood group O non-secretors are the most affected of the non-secretor blood types. One of the innate defenses against superficial infections by Candida species appears to be the ability of an individual to secrete the water-soluble form of his ABO blood group antigens into body fluids. The protective effect afforded by the secretor gene might be due to the ability of glycocompounds in the body fluids of secretors to inhibit adhesins (attachment lectins) on the surface of the yeast. In attachment studies, preincubation of blastospores with boiled secretor saliva significantly reduced their ability to bind to epithelial cells. ABH non secretor saliva did not reduce the binding and often enhanced the numbers of attached yeasts. (66,67) In one study, among individuals with Type II diabetes, 44% of ABH non-secretors were oral carriers of this yeast. (68)
Although non-secretors make up only about 26% of the population, they are significantly over represented among individuals with either oral or vaginal Candida infections, making up almost 50% of affected individuals. (69) The inability to secrete blood group antigens in saliva also appears to be a risk factor in the development of, or persistence of chronic hyperplastic Candidosis. In one study, the proportion of non-secretors of blood group antigens among patients with chronic hyperplastic Candidosis was 68%. (70)
Women with recurrent idiopathic vulvovaginal Candidiasis are much more likely to be ABH non-secretors. Combining both ABH non-secretor phenotype and absence of the Lewis gene Lewis (a- b-), the relative risk of chronic recurring vulvovaginal Candidiasis is between 2.41-4.39, depending on the analysis technique and control group. (71)
Oral carriage of Candida is also significantly associated with blood group O (p less than 0.001) and independently, with non-secretion of blood group antigens (p less than 0.001), with the trend towards carriage being greatest in group O non-secretors. (72)
ABH non-secretors appear to have an increase in the prevalence of a variety of autoimmune diseases including ankylosing spondylitis, reactive arthritis, psoriatic arthropathy, Sjogren’s syndrome, multiple sclerosis, and Grave’s disease. This susceptibility towards autoimmune problems appears to be most pronounced among Lewis (a-b-) phenotypes. Among individuals with spondyloarthropathies, non-secretors are reported to make up 47% of the patient population. In the subgroup of these patients suffering from ankylosing spondylitis, ABH non-secretors account for 49% of patients. Since the control population had a prevalence of non-secretors of 27% (consistent with the expected percent in the general population), it appears that in spondyloarthropathies in general, and ankylosing spondylitis specifically, non-secretors are dramatically over represented. (73,74)
Among individuals with primary Sjogren’s syndrome, Lewis blood group frequency differs from that of the general population, due mainly to an increased Lewis negative phenotype (Le (a-b-)) frequency. (75)
The inability to secrete the water soluble glycoprotein form of the ABO blood group antigens into saliva is significantly more common in patients with Graves’ disease than control subjects (40% versus 27%: p less than 0.025) but not among those with Hashimoto’s thyroiditis or spontaneous primary atrophic hypothyroidism.
ABH non-secretors with Grave’s disease were found to produce higher levels of antitubulin antibodies, while levels of other antibodies were similar to secretors. (76)
ABH Non-secretors are at an increased risk for development of celiac disease. One study reported that to 48% of patients with celiac disease were reported to be ABH non-secretors. (77) This appears to be especially true for the recessive Lewis (a-b-) phenotype. Evidence suggests an increased prevalence of complications and celiac-associated abnormalities is also found in the non-secreting and negative Lewis celiac patients. (78)
ABH secretors are significantly over represented among patients with influenza viruses A and B (55/64, 86%; p less than 0.025), rhinoviruses (63/72, 88%; p less than 0.01), respiratory syncytial virus (97/109, 89%; p less than 0.0005), and echoviruses (44/44, p less than 0.0005). Why this increased risk appears in secretors has not been clearly established. (79)
Among coal miners, asthma was significantly related to non-secretor phenotype. In this population, significantly lower lung function and higher likelihood of wheezing is especially prevalent among Lewis-negative or non-secretor subjects with blood group O. (80) Independent findings suggest that the ability to secrete ABH antigens might decrease the risk of COPD. Non-secretors have been found to have significantly greater impairment of forced expiration. ABH non secretors have lower mean values of forced expiratory volume in one second as a percentage of forced vital capacity (FEV1/FVC%) and a significantly larger proportion of them had aberrant values, defined as FEV1/FVC% less than 68. (81)
ABH non-secretor status also offers a slight increase risk for habitual snoring. (82)
NEOPLASIA AND MALIGNANCY
Secretor and Lewis Phenotypes and Tumor Markers
Accurately predicting the relevance of some tumor markers for diagnosis of cancer appears to be dependent on both secretor status and Lewis blood group. As an example, some researchers have suggested that taking into account aspects of Lewis and/or Secretor status in order to establish reference ranges might actually be a way to increase the clinical utility of the CA 19-9 tumor marker. (83)
There is a substantial difference in levels of this tumor marker are under the control of Secretor and Lewis genetics. Individuals having homozygous inactive Se alleles (se/se) and homozygous active Le alleles (Le/Le), exhibited the highest mean CA19-9 value. All of the Lewis negative individuals (Le (a- b-) consisting of a le/le genotype) had completely negative CA19-9 values, irrespective of the Segenotype.
On the other hand, Lewis negative individuals showed a higher mean DU-PAN-2 value than did the Le-positive individuals. Among patients with colorectal cancer, the Le-negative patients (le/le) with colorectal cancer showed undetectable CA19-9 values, i.e., less than 1.0 unit/ml, but many of them exhibited highly positive DU-PAN-2 values. In contrast, many of the Le-positive patients (Le/Le or Le/le) had positive CA19-9 values, whereas very few of them exhibited positive DU-PAN-2 values. (84)
None to very low levels
Table 4: CA19-9 and DU-PAN-9 expression in colorectal cancer correlated to Lewis subtype.
This implies that the CA19-9 measurement is not a useful tumor marker for Le-negative individuals, although DU-PAN-9 appears to be. Lewis negative individuals do not express any kinds of type 1 chain Lewis antigens (Lewis (a), Lewis (b), and secretory Lewis(a)) in their digestive organs. It is, therefore, not useful to measure the CA19-9 titer of the Lewis negative cancer patient. (85)
Preneoplastic Changes and Cancer
As a general rule, a higher intensity of oral disease is found among ABH non-secretors. So it is not surprising that when it comes to precancerous, or cancerous changes to tissue of the mouth and esophagus, ABH non-secretors seem to fair worse than ABH secretors. This oral disease susceptibility is reflected in the occurrence of epithelial dysplasia, for example, which is found almost exclusively in the non-secretor group. (86)
Barrett’s esophagus, a condition often preceding the development of esophageal cancer, and esophageal cancer also have a positive association with Lewis (a+b-) non-secretor phenotypes. (87)
The author wishes to thank Gregory Kelly, ND for his help in the preparation of this manuscript.
ABH SECRETOR STATUS TESTING
A single saliva-based test can be ordered from North American Pharmacal.
2009 Summer Street , Stamford CT 06905 USA
1. Bals R, Woeckel W, Welsch U. Use of antibodies directed against blood group substances and lectins together with glycosidase digestion to study the composition and cellular distribution of glycoproteins in the large human airways. J Anat 1997 Jan;190 ( Pt 1):73-84
2. Prakobphol A, Leffler H, Fisher SJ. The high-molecular-weight human mucin is the primary salivary carrier of ABH, Le(a), and Le(b) blood group antigens. Crit Rev Oral Biol Med 1993;4(3-4):325-33
3. Mohn JF, Owens NA, Plunkett RW. The inhibitory properties of group A and B non-secretor saliva. Immunol Commun 1981;10(2):101-26
4. Kogure T, Furukawa K. Enzymatic conversion of human group O red cells into Group B active cells by alpha-D-galactosyltransferases of sera and salivas from group B and its variant types. J Immunogenet 1976 Jun;3(3):147-54
5. Kapadia A, Feizi T, Jewell D, et al. Immunocytochemical studies of blood group A, H, I, and i antigens in gastric mucosae of infants with normal gastric histology and of patients with gastric carcinoma and chronic benign peptic ulceration. J Clin Pathol 1981 Mar;34(3):320-37
6. Triadou N, Audran E, Rousset M, et al. Relationship between the secretor status and the expression of ABH blood group antigenic determinants in human intestinal brush-border membrane hydrolases. Biochim Biophys Acta 1983 Dec 27;761(3):231-6
7. Domar U, Hirano K, Stigbrand T. Serum levels of human alkaline phosphatase isozymes in relation to blood groups. Clin Chim Acta 1991 Dec 16;203(2-3):305-13
8. Mehta NJ, Rege DV, Kulkarni MB. Total serum alkaline phosphatase (SAP) and serum cholesterol in relation to secretor status and blood groups in myocardial infarction patients. Indian Heart J 1989 Mar;41(2):82-85
9. Tibi L, Collier A, Patrick AW, et al. Plasma alkaline phosphatase isoenzymes in diabetes mellitus. Clin Chim Acta 1988 Oct 14;177(2):147-155
10. Agbedana EO, Yeldu MH. Serum total, heat and urea stable alkaline phosphatase activities in relation to ABO blood groups and secretor phenotypes. Afr J Med Med Sci 1996 Dec;25(4):327-9
11. Matsushita M, Irino T, Stigbrand T, et al. Changes in intestinal alkaline phosphatase isoforms in healthy subjects bearing the blood group secretor and non-secretor. Clin Chim Acta 1998 Sep 14;277(1):13-24
12. Hoskins LC, Boulding ET. Degradation of blood group antigens in human colon ecosystems. II. A gene interaction in man that affects the fecal population density of certain enteric bacteria. J Clin Invest 1976 Jan;57(1):74-82
13. Hoskins LC, Boulding ET. Degradation of blood group antigens in human colon ecosystems. I. In vitro production of ABH blood group-degrading enzymes by enteric bacteria. J Clin Invest 1976 Jan;57(1):63-73
14. Patrick AW, Collier A. An infectious aetiology of insulin-dependent diabetes mellitus? Role of the secretor status. FEMS Microbiol Immunol 1989 Jun;1(6-7):411-416
15. Peters WH, Gohler W. ABH-secretion and Lewis red cell groups in diabetic and normal subjects from Ethiopia. Exp Clin Endocrinol 1986 Nov;88(1):64-70
16. Melis C, Mercier P, Vague P, Vialettes B. Lewis antigen and diabetes. Rev Fr Transfus Immunohematol 1978 Sep;21(4):965-71
17. Eff C, Faber O, Deckert T. Persistent insulin secretion, assessed by plasma C-peptide estimation in long-term juvenile diabetics with a low insulin requirement. Diabetologia 1978 Sep;15(3):169-72
18. Hein HO, Sorensen H, Suadicani P, Gyntelberg F. The Lewis blood group–a new genetic marker of ischaemic heart disease. J Intern Med 1992 Dec;232(6):481-7
19. Ellison RC, Zhang Y, Myers RH, et al. Lewis blood group phenotype as an independent risk factor for coronary heart disease (the NHLBI Family Heart Study). Am J Cardiol 1999 Feb 1;83(3):345-8
20. .Zhiburt BB, Chepel’ AI, Serebrianaia NB, The Lewis antigen system as a marker of IHD risk. Ter Arkh 1997;69(1):29-31 [Article in Russian] Slavchev S, Tsoneva M, Zakhariev Z. The secretory type of persons who have survived a myocardial infarct. Vutr Boles 1989;28(2):31-4 [Article in Bulgarian]
21. Hein HO, Sorensen H, Suadicani P, Gyntelberg F. Alcohol intake, Lewis phenotypes and risk of ischemic heart disease. The Copenhagen Male Study. Ugeskr Laeger 1994 Feb 28;156(9):1297-302
22. Cruz-Coke R. Genetics and alcoholism. Neurobehav Toxicol Teratol 1983 Mar-Apr;5(2):179-80
23. Kojic T, Dojcinova A, Dojcinov D, et al. Possible genetic predisposition for alcohol addiction. Adv Exp Med Biol 1977;85A:7-24
24. Petit JM, Morvan Y, Viviani V, Vaillant G, Matejka G, Rohmer JF, Guignier F, Verges B, Brun JM. Related Articles Insulin resistance syndrome and Lewis phenotype in healthy men and women. Horm Metab Res. 1997 Apr;29(4):193-5
25. Hein HO, Sorensen H, Suadicani P, Gyntelberg F. Alcohol consumption, Lewis phenotypes, and risk of ischaemic heart disease. Lancet 1993 Feb 13;341(8842):392-6
26. Petit JM, Morvan Y, Mansuy-Collignon S, Viviani V, et al. Hypertriglyceridaemia and Lewis (A-B-) phenotype in non-insulin-dependent diabetic patients. Diabetes Metab 1997 Jun;23(3):202-4
27. Clausen JO, Hein HO, Suadicani P, et al. Lewis phenotypes and the insulin resistance syndrome in young healthy white men and women. Am J Hypertens 1995 Nov;8(11):1060-6
28. Viverge D, Grimmonprez L, Cassanas G, et al. Discriminant carbohydrate components of human milk according to donor secretor types. J Pediatr Gastroenterol Nutr 1990 Oct;11(3):365-70
29. Orstavik KH, Kornstad L, Reisner H, Berg K. Possible effect of secretor locus on plasma concentration of factor VIII and von Willebrand factor. Blood 1989 Mar;73(4):990-3
30. Wahlberg TB, Blomback M, Magnusson D. Influence of sex, blood group, secretor character, smoking habits, acetylsalicylic acid, oral contraceptives, fasting and general health state on blood coagulation variables in randomly selected young adults. Haemostasis 1984;14(4):312-9
31. Orstavik KH. Genetics of plasma concentration of von Willebrand factor. Folia Haematol Int Mag Klin Morphol Blutforsch 1990;117(4):527-31
32. Orstavik KH, Kornstad L, Reisner H, Berg K. Possible effect of secretor locus on plasma concentration of factor VIII and von Willebrand factor. Blood 1989 Mar;73(4):990-3
33. Green D, Jarrett O, Ruth KJ, Folsom AR, Liu K. Relationship among Lewis phenotype, clotting factors, and other cardiovascular risk factors in young adults. J Lab Clin Med 1995 Mar;125(3):334-339
34. Arneberg P, Kornstad L, Nordbo H, Gjermo P. Less dental caries among secretors than among non-secretors of blood group substance. Scand J Dent Res 1976 Nov;84(6):362-6
35. Al-Agidi SK, Shukri SM. Association between immunoglobulin levels and known genetic markers in an Iraqi population. Ann Hum Biol 1982 Nov-Dec;9(6):565-9
36. Shinebaum R. ABO blood group and secretor status in the spondyloarthropathies. FEMS Microbiol Immunol 1989 Jun;1(6-7):389-95
37. Grundbacher FJ. Immunoglobulins, secretor status, and the incidence of rheumatic fever and rheumatic heart disease. Hum Hered. 1972;22(4):399-404.
38. Grundbacher FJ. Genetic aspects of selective immunoglobulin A deficiency. J Med Genet. 1972 Sep;9(3):344-7.
39. Dube VE, Tanaka M, Chmiel J, Anderson B. Effect of ABO group, secretor status and sex on cold hemagglutinins in normal adults. Vox Sang 1984;46(2):75-9
40. Blackwell CC, Weir DM, Patrick AW, Collier A, Clarke BF. Secretor state and complement levels (C3 and C4) in insulin dependent diabetes mellitus. Diabetes Res 1988 Nov;9(3):117-9
41. Tandon OP, Bhatia S, Tripathi RL, Sharma KN. Phagocytic response of leucocytes in secretors and non-secretors of ABH (O) blood group substances. Indian J Physiol Pharmacol 1979 Oct-Dec;23(4):321-4
42. Odeigah PG. Influence of blood group and secretor genes on susceptibility to duodenal ulcer. East Afr Med J 1990 Jul;67(7):487-500
43. Suadicani P, Hein HO, Gyntelberg F. Genetic and life-style determinants of peptic ulcer. A study of 3387 men aged 54 to 74 years: The Copenhagen Male Study. Scand J Gastroenterol 1999 Jan;34(1):12-7
44. Hein HO, Suadicani P, Gyntelberg F. Genetic markers for stomach ulcer. A study of 3,387 men aged 54-74 years from The Copenhagen Male Study. Ugeskr Laeger 1998 Aug 24;160(35):5045-46
45. Dickey W, Collins JS, Watson RG, et al. Secretor status and Helicobacter pylori infection are independent risk factors for gastroduodenal disease. Gut 1993 Mar;34(3):351-3
46. Sumii K, Inbe A, Uemura N, et al. Multiplicative effect of hyperpepsinogenemia I and non-secretor status on the risk of duodenal ulcer in siblings. Gastroenterol Jpn 1990 Apr;25(2):157-61
47. Dickey W, Collins JS, Watson RG, et al. Secretor status and Helicobacter pylori infection are independent risk factors for gastroduodenal disease. Gut 1993 Mar;34(3):351-3
48. Oberhuber G, Kranz A, Dejaco C, et al. Blood groups Lewis(b) and ABH expression in gastric mucosa: lack of inter-relation with Helicobacter pylori colonisation and occurrence of gastric MALT lymphoma. Gut 1997 Jul;41(1):37-42
49. Su B, Hellstrom PM, Rubio C, et al. Type I Helicobacter pylori shows Lewis(b)-independent adherence to gastric cells requiring de novo protein synthesis in both host and bacteria. J Infect Dis 1998 Nov;178(5):1379-90
50. Alkout AM, Blackwell CC, Weir DM, et al. Isolation of a cell surface component of Helicobacter pylori that binds H type 2, Lewis(a), and Lewis(b) antigens. Gastroenterology 1997 Apr;112(4):1179-87
51. Klaamas K, Kurtenkov O, Ellamaa M, Wadstrom T. The Helicobacter pylori seroprevalence in blood donors related to Lewis (a,b) histo-blood group phenotype. Eur J Gastroenterol Hepatol 1997 Apr;9(4):367-70
52. Mentis A, Blackwell CC, Weir DM, et al. ABO blood group, secretor status and detection of Helicobacter pylori among patients with gastric or duodenal ulcers. Epidemiol Infect 1991 Apr;106(2):221-9
53. Sheinfeld J, Schaeffer AJ, Cordon-Cardo C, et al. Association of the Lewis blood-group phenotype with recurrent urinary tract infections in women. N Engl J Med 1989 Mar 23;320(12):773-7
54. Similar findings by other researchers support this over representation of recurrent UTI among non-secretors both in women and children.
55. May SJ, Blackwell CC, Brettle RP, MacCallum CJ, Weir DM. Non-secretion of ABO blood group antigens: a host factor predisposing to recurrent urinary tract infections and renal scarring. FEMS Microbiol Immunol 1989 Jun;1(6-7):383-7
56. A, Nudelman E, Clausen H, et al. Binding of uropathogenic Escherichia coli R45 to glycolipids extracted from vaginal epithelial cells is dependent on histo-blood group secretor status. J Clin Invest 1992 Sep;90(3):965-72
57. Jantausch BA, Criss VR, O’Donnell R, et al. Association of Lewis blood group phenotypes with urinary tract infection in children. J Pediatr 1994 Jun;124(6):863-8
58. Kinane DF, Blackwell CC, Brettle RP, et al. ABO blood group, secretor state, and susceptibility to recurrent urinary tract infection in women. Br Med J (Clin Res Ed) 1982 Jul 3;285(6334):7-9
59. May SJ, Blackwell CC, Brettle RP, MacCallum CJ, Weir DM. Non-secretion of ABO blood group antigens: a host factor predisposing to recurrent urinary tract infections and renal scarring. FEMS Microbiol Immunol 1989 Jun;1(6-7):383-7
60. Lomberg H, Hellstrom M, Jodal U, Svanborg Eden C. Secretor state and renal scarring in girls with recurrent pyelonephritis. FEMS Microbiol Immunol 1989 Jun;1(6-7):371-5
61. Lomberg H, de Man P, Svanborg Eden C. Bacterial and host determinants of renal scarring. APMIS 1989 Mar;97(3):193-9
62. Jacobson SH, Lomberg H. Overrepresentation of blood group non-secretors in adults with renal scarring. Scand J Urol Nephrol 1990;24(2):145-50
63. Lomberg H, Jodal U, Leffler H, et al. Blood group non-secretors have an increased inflammatory response to urinary tract infection. Scand J Infect Dis 1992;24(1):77-83
64. Blackwell CC, Weir DM, James VS, et al. Secretor status, smoking and carriage of Neisseria meningitidis. Epidemiol Infect 1990 Apr;104(2):203-9
65. Zorgani AA, Stewart J, Blackwell CC, Elton RA, Weir DM. Inhibitory effect of saliva from secretors and non-secretors on binding of meningococci to epithelial cells. FEMS Immunol Med Microbiol 1994 Aug;9(2):135-42
66. Thom SM, Blackwell CC, MacCallum CJ, et al. Non-secretion of blood group antigens and susceptibility to infection by Candida species. FEMS Microbiol Immunol 1989 Jun;1(6-7):401-5
67. Ben-Aryeh H, Blumfield E, Szargel R, et al. Oral Candida carriage and blood group antigen secretor status. Mycoses 1995 Sep-Oct;38(9-10):355-8
68. Blackwell CC, Aly FZ, James VS, et al. Blood group, secretor status and oral carriage of yeasts among patients with diabetes mellitus. Diabetes Res 1989 Nov;12(3):101-4
69. Thom SM, Blackwell CC, MacCallum CJ, et al. Non-secretion of blood group antigens and susceptibility to infection by Candida species. FEMS Microbiol Immunol 1989 Jun;1(6-7):401-
70. Lamey PJ, Darwazeh AM, Muirhead J, et al. Chronic hyperplastic candidosis and secretor status. J Oral Pathol Med 1991 Feb;20(2):64-7
71. Chaim W, Foxman B, Sobel JD. Association of recurrent vaginal candidiasis and secretory ABO and Lewis phenotype. J Infect Dis 1997 Sep;176(3):828-30
72. Burford-Mason AP, Weber JC, Willoughby JM. Oral carriage of Candida albicans, ABO blood group and secretor status in healthy subjects. J Med Vet Mycol 1988 Feb;26(1):49-56
73. Shinebaum R. ABO blood group and secretor status in the spondyloarthropathies. FEMS Microbiol Immunol 1989 Jun;1(6-7):389-95
74. Shinebaum R, Blackwell CC, Forster PJ, et al. Non-secretion of ABO blood group antigens as a host susceptibility factor in the spondyloarthropathies. Br Med J (Clin Res Ed) 1987 Jan 24;294(6566):208-10
75. Manthorpe R, Staub Nielsen L, et al. Lewis blood type frequency in patients with primary Sjogren’s syndrome. A prospective study including analyses for A1A2BO, Secretor, MNSs, P, Duffy, Kell, Lutheran and rhesus blood groups. Scand J Rheumatol 1985;14(2):159-62
76. Toft AD, Blackwell CC, Saadi AT, et al. Secretor status and infection in patients with Graves’ disease. Autoimmunity 1990;7(4):279-89
77. Dickey W, Wylie JD, Collins JS, et al. Lewis phenotype, secretor status, and coeliac disease. Gut 1994 Jun;35(6):769-70
78. Heneghan MA, Kearns M, Goulding J, et al. Secretor status and human leucocyte antigens in coeliac disease. Scand J Gastroenterol 1996 Oct;31(10):973-6
79. Raza MW, Blackwell CC, Molyneaux P, et al. Association between secretor status and respiratory viral illness. BMJ 1991 Oct 5;303(6806):815-8
80. Kauffmann F, Frette C, Pham QT, et al. Associations of blood group-related antigens to FEV1, wheezing, and asthma. Am J Respir Crit Care Med 1996 Jan;153(1):76-82
81. Cohen BH, Bias WB, Chase GA, et al. Is ABH nonsecretor status a risk factor for obstructive lung disease. Am J Epidemiol 1980;3:285-91
82. Jennum P, Hein HO, Suadicani P, et al. Snoring, family history, and genetic markers in men. The Copenhagen Male Study. Chest 1995 May;107(5):1289-93
83. Vestergaard EM, Hein HO, Meyer H, et al. Reference values and biological variation for tumor marker CA 19-9 in serum for different Lewis and secretor genotypes and evaluation of secretor and Lewis genotyping in a Caucasian population. Clin Chem 1999 Jan;45(1):54-61
84. Narimatsu H, Iwasaki H, Nakayama F, et al. Lewis and secretor gene dosages affect CA19-9 and DU-PAN-2 serum levels in normal individuals and colorectal cancer patients. Cancer Res 1998 Feb 1;58(3):512-8
85. Narimatsu H. Molecular biology of Lewis antigens–histo-blood type antigens and sialyl Lewis antigens as tumor associated antigens. Nippon Geka Gakkai Zasshi 1996 Feb;97(2):115-22 [Article in Japanese]
86. Vidas I, Delajlija M, Temmer-Vuksan B, et al. Examining the secretor status in the saliva of patients with oral pre-cancerous lesions. J Oral Rehabil 1999 Feb;26(2):177-82
87. Torrado J, Ruiz B, Garay J, et al. Blood-group phenotypes, sulfomucins, and Helicobacter pylori in Barrett’s esophagus. Am J Surg Pathol 1997 Sep;21(9):1023-9
88. Stolbach LL, Krant MJ, Fishman WH. Intestinal alkaline phosphatase in chylous effusion: role of ABO blood group and secretor status. Enzymologia 1972 Jun 30;42(6):431-8
89. Walker BA, Eze LC, Tweedie MC, Evans DA. The influence of ABO blood groups, secretor status and fat ingestion on serum alkaline phosphatase. Clin Chim Acta 1971 Dec;35(2):433-44
90. Bayer PM, Hotschek H, Knoth E Intestinal alkaline phosphatase and the ABO blood group system–a new aspect. Clin Chim Acta 1980 Nov 20;108(1):81-87