Interaction between drugs and foods

Published on Editorials  

One of the main factors hampering the development of drugs at the up-start of the scientific therapy era was the fear that it would be very difficult to control both the huge number of variables surrounding the pathologies, and the effects of the drugs themselves. However, so called «clinical observations in the field» made it possible to closely define, describe, and quantify the patients’ responses, with regards to both preliminary studies and population studies. When one drug alone is used, it is fairly easy to monitor its metabolic transformations since its primary effects or any side effects can be observed almost univocally, at least as far as its basic trends are concerned. Although it is true that identifying possible damage is not always easy, especially when the response is not what we expect, the benefit is far greater than the risk. The situation becomes even more complicated when two or more drugs are used simultaneously. Combined therapy with two or more drugs is far from rare, but unfortunately is often mandatory in order to cure or to improve the patient's health, as for example with hypertension, cardiac insufficiency, and tumor syndromes. In cases like these, facing the problems concerning the possible interactions of the drugs becomes a priority issue, since the overall outcome may be an increase or a decrease in the effects of one or more of the drugs. In other words, the effects may range from total inefficacy to full-blown toxicity. Some combined therapy schedules are so well standardized that the health risks are very low. While in other and far more frequent therapy schedules, combined treatments are fortuitous therefore making it important to identify and ascribe their interactions. Taking all these factors into consideration, together with the lack of feasible monitoring, the probability of mistakes or of possible damage caused by combined therapy is very high, and in view of the increasingly large number of new drugs that are placed on the market every year, the complexity and seriousness of some choices may only become more difficult. Furthermore, another problem which is closely connected to the previously mentioned issue is now arising, even though, to date, it has been limited to the academic field.
Most people consider the concept of diet-pharmacology as a purely scientific curiosity and it has only recently been acknowledged that some nutrients may act like medicines, which themselves may have some of the characteristics of foods. Until now, only clinical pharmacologists have been aware that we cannot understimate either the effects that drugs have on food intake, or the dietary factors affecting the biological availability of a drug. The interaction between foods and drugs may be of the physical-chemical type (e.g. absorption, formation of complex ions/salts, precipitation), physiological (such as dietary variables affecting the bio-availability of a drug), or even physiopathologic (toxic effects of the drug affecting food absorption). The physical-chemical interaction usually leads to a decrease in the absorption of one or both factors. This is the case when tetracycline is administered within two hours before/after intaking food or other supplements containing calcium.(1) The physiologic interactions may go in one of two directions, i.e., weak or strong absorption of a medicine or of a nutrient. These effects are usually reversible and transient, but they may sometimes strongly delay the pharmacological action or cause greater toxicity or even drastic malabsorption of a nutrient with striking consequences in cases of prolonged therapy. Delayed absorption of a drug may sometimes be useful and desirable, mostly because there will be fewer side effects. However when the interaction speeds up the uptake, toxicity may become acute. For example, when digoxin is taken, hemicellulose dietary fibers slow down drug absorption (likely because this fiber increases the amount of time needed for gastric emptying).(2) Something similar occurs when nifedipine, a known calcium channel blocker, is administered together with a low-fat meal.(3) In this case the effect is favourable for patients who are easily affected by vasogenic reddening caused by this type of drug. Conversely, a high-fat meal will speed up absorption of preparations with theophylline, which increases toxicity, thus causing headache, tachycardia, and agitation.(4) It is also well known that diarrhea caused by laxative abuse may lead to very severe hypokalemia. Physiopathologic interactions include situations in which drug toxicity blocks some biosynthesis mechanisms, thus leading the organism to inactivate or to fail to use the nutrients. We know, for example, of the blocking action that methotrexate, trimetrexate, and sulfasalazine, has on folates.(5) Physical-chemical interactions may be classified according to their effects on: absorption, solubility, precipitation, chelation, ion-exchange, and photo activation. On the other hand, physiologic interference affecting the gastro-intestinal function may cause hyper- or hypophagia as well as electrolytic imbalance.
Physiopathologic damage is characterized by toxicity in various organs: liver, kidney, nervous system, embryotoxicity, and hemolysis. Of course all these mechanisms may be more or less harmful depending on the patient's illness, i.e. cardiopathy, diabetes, and so on. We must keep in mind that not only do patients take drugs to treat their illnesses, but they also take other drugs against the side-effects of the same illness. Indeed co-morbidity is another cause of imbalance. Severe iron deficiency anemia is not rare in cases of maturity-onset diabetes observed with colon neuropathy. In such patients, treatment with non steroid anti-inflammatory drugs, for example when osteopathy is present, may well lead to ulceration of the tract, causing blood loss and hence anemia, which is often worsened by concomitant anti-coagulant therapy, or high alcohol intake. We see examples almost daily of incorrect combinations of drugs, or of the use of medicines whose effects might be reached by other types of therapy which are less harmful to existing lesions. An example of this is the choice of thiazide-based diuretics to treat hypertension. Since these substances favor hyperglycemia, they may unbalance an already existing diabetes, or precipitate pre diabetic latency. A slight rise in triglycerides is usually observed, together with increased serum cholesterol levels. In cases like these it would be easy to change the drug. Investigating the area of «wrong associations», even on a superficial basis, broadens the whole spectrum. How often are iron salts administered without knowing that the presence of antacids, caffeine, cimetidine, and tetracyclines (doxycyclines excepted) in the bowel will block these salts thus reducing their bioavailability? These salts should be administered at least two-three hours before/after meals. Drugs which inhibit dihydro-folate-reductase (methotrexate, trimetoprim), and those that prevent or reduce the absorption or storage of folic acid ( oral contraceptives, antiepileptic drugs) may lead to a lack of folic acid and therefore, megaloblastic anemia.
Alcohol increases the effects of dicumarol based acecoumarol, an anti-coagulant, but only in case of acute poisoning, whereas in case of chronic abuse the opposite effect is observed. The same thing happens with Chlorpropamide. Its effect, which stimulates the production of insulin by the body, increases in the former case, while it decreases in the latter. The recommended daily dose of calcium (800-1200 mg) may become drastically insufficient if we do not take into account that foods which are rich in oxalates (cacao, spinach, soya beans), phosphates (natural rice, bran, flour), and phytates (wholemeal bread) reduce its absorption, while calcium absorption is increased by citrates, lactose and, - as is well known – by vitamin D. The consumption of H-chain fatty acids promotes the absorption of liposoluble vitamins, but may increase calcium loss through the feces.(6) The current and understandable trend towards reducing dietary fat intake (obesity is on the rise world-wide) has led to the synthesis of replacement products (olestra for example), but these reduce the absorption of liposoluble vitamins and of some drugs.(7) Generally speaking, fibers (especially insoluble fibers) may reduce the bioavailability of iron, zinc, calcium, and magnesium,(8) although reducing the fiber content in foods and increasing the intake of proteins leads to similar results. It has long been known that the type of diet affects the composition of the intestinal microflora (low meat consumption and high complex carbohydrate consumption increase aerobic bacteria while reducing the number of a given type of anaerobic bacteria), thus conditioning metabolic activity and facilitating a possible rebound on the host.(9)
Another equally interesting, though subtle and potentially unsettling problem arises along with this subject, i.e., when the nutritive principle acts like a drug, as has often been noticed in vitamins and the like. In 1954 Miescher had already highlighted the inhibiting effect of high doses of vitamin A on keratinization. Doses ranging between 200,000 and 400,000 IU/die caused regression of the ichthyosis syndrome within two – three months. However such doses (500,0000 UI/die) administered to children caused abnormal skeletal development and hemorrhages. They reported similar observations for vitamin D3 (2,500-15,000 micrograms/day), the onset of hypocalcemia, and changes in kidney function together with other signs. Another example is Niacin, whose vasodilating, and triglyceride and LDL cholesterol lowering effects have favoured its use, but doses > 300 mg/day cause reddening, dryness of the skin, nausea, diarrhea and in some cases, acantosi nigricans type skin alterations. Unfortunately, this list could be much longer, but it is of course long enough to suggest that the issue is very complex and wide ranging. The authors have intentionally left out the dietary factors affecting drug bioavailability (time of gastric emptying, etc.). Treatment of this subject is certainly far from complete and in some ways it is not even exhaustive, though we may well expect that if we should come into direct or indirect contact with this subject, we would be less superficial and at least stimulated to consider the matter with the benefit of the doubt.


  1. Hansten, P.D.: Drug Interactions. 5 Ed. Philadelphia, Lea&Febiger, 1985.
  2. Nordstrom, M., Melander,A., Robertson, E. et al.: Drug-Nutrient Interact., 5:67-69, 1987.
  3. Reitberg, D.P., Lowe, S.J., Quercia,G.T. et al.: Clin. Pharmacol. Ther., 42:72-75, 1987.
  4. Vaughan, L., Milavetz, G., Hil, M., et al.: Drug Intell. Clin. Pharm., 18:510-513, 1984.
  5. Roe, D. A.: Drug folate interrelationships: Historical aspects and current concerns. In Folic Acid Metboslism in Health and Disease. Edited by M.F. Picciano, E.L.R. Stokstad, J.F. Gregory. New York, Wiley-Liss, 1990.
  6. Silk, D.B.A., Sawson, A.M.: International Reviews of Physiology: Gastrointestinal Physiology III. Vol. 19. Edited by R. H. Crane. Baltimore, Univeristy Park Press, 1979, pp.151-204.
  7. Jones, D.Y., Miller, K.W., Koonsvitsky, B.P., et al.: Am.J.Clin.Nutr., 53:1281-1287, 1991.
  8. Kelsay, J.: In “Dietary Fiber in Health and Disease”. Edited by G.V. Vahouny and D. Kritchevsky. New York, Plenum, 1982, pp. 91-103.
  9. Maier, B.R., et al.: Am.J.Clin.Nutr., 27:1470-1474, 1974