A review of the literature associates the ABO blood groups with a wide assortment of pathologic conditions, especially gastrointestinal and cardiovascular pathology.
ABO blood groups, ABH antigen system, achlorhydria, carcinoma peptic ulcer, gastric carcinoma, ischemic heart disease, serum cholesterol alkaline phosphatase isoenzymes, toxemia, plasma factor VIII levels
Alexander () was probably the first researcher to correlate the blood groups to any type of pathology, giving the following frequencies for fifty patients suffering from various types of carcinoma : 0: 14, A:14, B:16, AB:6 contrasting these with 175 results from 50 normals, 50 tuberculars, 50 syphilis and 25 tetany patients, concluding that groups B and AB were especially prone to cancer; a conclusion echoed by later workers Mithra.() in 1933 and Pautienis () in 1937 who both noticed a greater propensity of groups A and AB to develop cancer.
Mayo and Ferguson () summarizing the available literature were led to conclude that the most common finding amongst the investigators seemed to be that group AB is elevated in cancer populations. A corroborative finding with trophoblastic neoplasia () links group AB to rapidly progressing choriocarcinomas which are highly resistant to treatment, the highest incidence associated with gravida of group A partnered to group O mates.
Stimulated by Alexander's paper, Buchanan and Higley () analyzed the results of 2446 patients grouped at the Mayo clinic.
|Pernicious anemia ||189||202||46||20|
|Fibroid Uterus ||54||64||13||6|
Looking at the percentage frequencies for the groups, they concluded that:
1. There is no relationship between blood type and any malignancy as suggested by Alexander.
2. There is no relationship between the blood groups and any disease in which there is sufficient data to justify a conclusion.
3. Nationality should be taken into account in the presentation of statistical studies on blood groups.
However, as Fraser Roberts illustrated in his critique () by being exceedingly conscious of racial difference between the blood types the evidence requiring heterogeneity is considerably weakened. Nevertheless, it was a shame that after looking at the figures that they did not conclude that group O was in excess with ulcers and group A with pernicious anemia. Had they done so others would have immediately made further studies on these diseases and two associations would have been discovered. Unfortunately their report served to sidetrack interest in the blood groups for a considerable period of time.
The remaining literature on carcinoma encompasses the significant through enigmatic. A number of studies were done on various cancers, with leukemia (, carcinoma of the bronchus (,,),and rectal carcinoma () showing a no significant associations; breast cancer having a suggested association (); pituitary adenoma () and pancreatic carcinoma () having significant associations and uterine carcinoma (,), salivary gland tumors() having highly significant association.
The association of group A with higher incidences of gastric carcinoma, first demonstrated by Aird, Bentall and Fraser Roberts in 1953 signaled the resurgence of clinical interest in the blood groups during the late 1950's and. 1960's. Their combining of newer, more powerful analytic tools with the large populations required for this sort of analysis foreshadowed renewed interest in an association between group O and peptic ulcer, first postulated by Ugelli in 1937.() In a paper published the following year, Aird et al () demonstrated a "strikingly high" increase (17.9%) of group O amongst ulcer populations compared to controls, with a correspondingly lower incidence of the other groups. This has been extensively corroborated by various centers (,,). Reports linking group O to decreased Factor VIII concentration () would seem to tie into these findings.
The associations between group A and gastric carcinoma and group O with peptic and duodenal ulcers seem based upon variations in physiologic parameters particular to each group. The previously mentioned association of group A with pernicious anemia () can now be regarded as significant with considerable confidence, as the findings appear to relate appropriately to reports from m India () documenting differences between group A and O in acid secretion in response to test meal. Sievers (), using a tubeless test meal technique found achlorhydria significantly more common amongst group A individuals with GI problems. Studies linking achlorhydria and pernicious anemia to gastric carcinoma (,) seem to indicate a distinct "functional milieu" between the blood types, especially regarding the gastrointestinal tract, a prime depository for ABH antigens.
One fascinating finding involves intestinal alkaline phosphatase, an isoenzyme shown to be related to the concentration of lymph triglycerides (); rising in response to fat or fatty acids in a meal. Arsfors () studied both fast and slow moving electrophoretic patterns of this enzyme and found the variations to be strongly under genetic control and definitely linked to the blood groups. These zones are relatively common in secretors but are infrequently seen in non secretors of any group. In secretor populations up O and B show the zone very much more frequently than group A with AB as an intermediary().
This physiologic variant appears to be causal regarding the next two AB0 related associations: hypercholesterolemia and ischemic heart disease. Langman() showed an inverse correlation of genetic factors associated with intestinal alkaline phosphatase and serum cholesterol and suggested that these intestinal factors may in part effect serum cholesterol levels. Oliver () in a survey of 6000 found that men belonging to groups O or B had a lower mean cholesterol level than group A, a finding that "emphasizes the genetic pathogenesis of ischemic heart disease".
The Framington Heart Study () typed 4125 survivors of an original cohort of 5209, recognizing 273 cardiac events. The most noteworthy finding was with men aged 39-72 where a lower incidence of non fatal coronary heart disease was observed with group O as compared with group A (x2=5.80, P less than 0.02), a finding corroborated in studies of 792 () and 353 (). These findings appear to correlate with studies demonstrating increased incidence of venous thromboembolism in group A populations.()
The evidence for portal cirrhosis, hyperthyroidism and diabetes is less strong and justifies no more than a provisional conclusion serving as a pointer towards profitable work in the future. Hyperthyroidism has been reported as being in excess in group O and D'Adamo reports hypothyroidism as being a condition common in group A (). Portal cirrhosis was found to have a highly significant association with group A, but the study numbered only 111. Diabetes has been studied extensively (,,) with a total cohort of 2150. Group A was found to be increased, however most reporters found the excess limited only to males whilst one researcher reported an association limited only to females.
Although I wrote this paper over 20 years ago, I think it should still be read by any clinician who doubts the widespread activity of ABO antigens outside the blood stream and their consistent association with disorders of the digestive tract. -PD
- D'Adamo, P. Submitted in partial fulfillment of Rounds/ Journal Club course requirement, Fall 1981, John Bastyr College of Naturopathic Medicine, Seattle WA