Summary. In a survey of 46 randomly selected diabetic patients on biguanide therapy, 30% had malabsorption of vitamin B12. Withdrawal of the drug resulted in normal absorption in only half of those with malabsorption. In most patients with persistent malabsorption, the results of absorption tests with exogenous intrinsic factor suggested the diagnosis of coincidental intrinsic factor deficiency. Further considerations, however, led to the concept that biguanides can induce malabsorption by two different mechanisms. One of these is temporary and unrelated to intrinsic factor secretion and the other is permanent and mediated by depression of intrinsic factor secretion.
Key-words:Biguanides, phenformin, metformin, vitamin B12 , intrinsic factor deficiency, malabsorption.Malabsorption of vitamin B12 is a well-known complication of biguanide therapy. Any disease causing malabsorption may, however, occur in diabetic patients and an interesting feature of studies of vitamin B12 absorption in diabetic patients on biguanides is the apparent rarity of such conditions. As we had formed an impression that anaemias not related to biguanide therapy occurred more often than might be expected, we decided to test this by a study of the absorption of vitamin Ba2 by diabetic patients on biguanides and extended this to the effects of metformin on the gastric secretory capacity of normal subjects.
Subjects and Methods
SubjectsForty-six diabetic patients on biguanide therapy were selected randomly for a survey of the capacity to absorb vitamin B12. Thirty-eight were female (aged 40-72 years, average 58 years) and eight were male (aged 45-72 years, average 59 years). The ratio of males to females is representative of the sex distribution of obese non-insulin-dependent diabetic patients attending the clinic. The average age of onset of diabetes was 50years in males (range 38-67years) and 51 years in females (range 32-75 years).Twenty-eight patients (24 males, four females) were studied on metformin 13 (three males, ten females) on phenformin and five (one male, four females) while on both drugs at different times. The dose of metformin ranged from i to 3 g daily, and the duration of treatment at the time of study was from 1 to 6 years; the dose of phenformin was 50 mg twice daily and the duration of treatment ranged between 3 months and 6 years.
MethodsThe capacity to absorb vitamin B~2 was measured with a scanning type whole body monitor utilising six sodium iodide detectors each 15.2 cm diameter and 10.2 cm thick housed in a custom built room with 15 cm thick steel walls. The statistical accuracy of a result is + 5% (i.e.: 5 + 5%, 20 _+ 5%, etc.) and the mean coefficient of variation between results is 26%. A control series absorbed between 26% and 89% (mean + SD: 50.6 -!-_ 14.7%) of 1 Ixg radioactive cyanocobalamin and for routine purposes absorption of < 30% is regarded as subnormal. After preliminary scanning, oral doses of 1 gg, 0.25 j_tCi, 57Co or 58Co cyanocobalamin were given after a t2-h fast which was continued for a further...