Interesting, but it is currently unknown why this occurs. From Science Daily:
People with blood groups A, B and AB at higher risk of type 2 diabetes than group O
A study of more than 80,000 women has uncovered different risks of developing type 2 diabetes associated with different blood groups, with the biggest difference a 35 percent increased risk of type 2 diabetes found in those with group B, Rhesus factor positive blood compared with the universal donor group O, Rhesus factor negative.
While previous studies have investigated the links between blood group and stroke (finding an increased risk for group AB versus group O), those on blood groups and their link with diabetes have been small and thus underpowered and unable to provide definitive results. In this new research, Fagherazzi and colleagues took data from 82,104 women from the large prospective E3N cohort in France followed between 1990 and 2008. The objective of this study was to evaluate the relationship of ABO blood type (A, B, AB and O), Rhesus factor (positive or negative) and a combination of the two (ABO×Rhesus) with type 2 diabetes (T2D).
The results showed that, compared with women with group O blood, women with group A were 10% more likely to develop T2D, and those with group B 21% more likely (both statistically significant). The AB group was 17% more likely to develop T2D, but this result was not statistically significant. When looking solely at R+ versus R- women, neither group was at increased risk of developing T2D compared with the other.
The authors then combined ABO group and Rhesus group and compared each possible combination with O negative (O-), which is known as the universal donor group because since it contains none of the A. B, or Rhesus antigens, blood from people in this group can be successfully donated to any other group (without rejection).Compared with O- women, the highest increased risk of developing T2D was found in B+ (35% increased risk), followed by AB+ (26%), A- (22%), A+ (17%). The results for O-, B-, and AB- groups were not statistically significant.