International Agency for Research on Cancer (IARC) - Summaries & Evaluations

(Group 3)

For definition of Groups, see Preamble Evaluation.

VOL.: 51 (1991) (p. 291)

CAS No.: 58-08-2
Chem. Abstr. Name: 3,7-Dihydro-1,3,7-trimethyl-1H-purine-2,6-dione

5. Summary of Data Reported and Evaluation

5.1 Exposure data

Caffeine is a methylxanthine, which occurs naturally in more than 60 plant species throughout the world. It is prepared on an industrial scale by methylation of theobromine.

Global per-caput consumption of caffeine from all sources was estimated to be 70 mg per day in 1981-82.

Caffeine is consumed in beverages such as coffee, tea and mate and in soft drinks to which caffeine is added. Coffee is the main source of dietary caffeine consumption. The caffeine content of beverages varies widely. Caffeine is also used in numerous prescription and non-prescription pharmaceutical preparations.

5.2 Experimental carcinogenicity data

Caffeine was tested for carcinogenicity in five studies in rats by oral administration. In two of these studies, no significant difference in the incidence of tumours at any site was found. The other three studies were found to be inadequate for evaluation.

Studies on oral and intraperitoneal administration of caffeine to mice were found to be inadequate for evaluation.

In one study, decaffeinated coffee to which caffeine was added was tested by oral administration to rats; overall, no increase in tumours at any site was observed as compared to appropriate controls.

Administration of caffeine in combination with known carcinogens resulted in decreased incidences of lung tumours in mice treated with urethane, of mammary tumours in rats treated with diethylstilboestrol and of skin tumours in mice treated with either ultra-violet light or cigarette-smoke condensate. Caffeine did not influence the incidence of bladder tumours induced in rats by N-nitroso-N-butyl(4-hydroxybutyl)amine in three experiments or of pancreatic tumours induced in rats by 4-hydroxyaminoquinoline-1-oxide in another study.

5.3 Human carcinogenicity data

A cohort study with a short follow-up period showed no association between caffeine consumption and mortality from cancers at all sites, although there were few deaths on which to base an analysis.

Three case-control studies of breast cancer in which an attempt was made to measure methylxanthine intake showed no association. A slight increase in risk was seen in premenopausal women in one study, but in general the relative risks were below unity.

One case-control study of bladder cancer showed a weak association with caffeine consumption.

Caffeine and coffee consumption are highly correlated in most of the populations studied; thus, it is very difficult to separate the two exposures in epidemiological studies. It was therefore not possible to evaluate adequately the effect of caffeine per se.

5.4 Other relevant data

Caffeine intake from pharmaceutical sources has not been related to teratogenic effects in humans. High levels of either coffee or caffeine consumption were related to an increased frequency of low birthweight.

Quantitative and qualitative differences in the metabolism of caffeine are seen between humans and experimental animals.

On the basis of the available evidence, caffeine consumed in moderate amounts does not cause any persistent increase in blood pressure in normotensive subjects. Whether caffeine consumed in amounts present in coffee or tea causes cardiac arrhythmias in healthy subjects or in patients with heart disease remains an open question.

Caffeine has been shown to cause adverse reproductive and developmental effects in mice, rats, rabbits and monkeys. Testicular atrophy was observed at high dose levels in rats. Reproductive studies in mice showed no effect on pregnancy but there was a decrease in litter size at birth. Teratogenic effects were usually associated with high, single, daily doses that were also associated with other signs of maternal toxicity. High daily levels given as divided doses were less toxic to the conceptus that when given as a single dose. Reduced fetal body weight was observed in rats. A reversible delay in ossification of the sternum was observed in rats at a relative low dose given by gavage. With administration in drinking-water, similar effects were seen, but at higher doses.

One epidemiological study revealed no effect of caffeine (in coffee-drinking subjects) on the sex ratio of their children. In lymphocytes of normal, caffeine-exposed people, chromosomal aberrations were not observed. An increased frequency of micronucleated blood cells was observed in otherwise healthy splenectomized people exposed to caffeine. Urine of caffeine-exposed persons was not mutagenic to Salmonella typhimurium.

Although it has been suggested that caffeine may induce gene mutations in mammals and man, direct evidence in vivo is limited and the indirect evidence is largely based on extrapolation from results in lower organisms, in which there is no doubt about the mutagenic action of caffeine, and from cultured mammalian cells, in which caffeine is clastogenic at high concentrations.

Overall, caffeine affects photoreactivation, excision repair and postreplication repair. The antagonistic effect of caffeine on mutations has induced by ultra-violet radiation has been explained on the basis of inhibition of an error-prone, postreplicative, recombination repair process. Caffeine can modulate the effects of xenobiotics by acting on (i) cytochrome P450, (ii) cAMP metabolism, (iii) DNA metabolism, chromatin structure and function and (iv) nucleotide pools.

5.5 Evaluation

There is inadequate evidence for the carcinogenicity in humans of caffeine.

There is inadequate evidence for the carcinogenicity of caffeine in experimental animals.

Overall evaluation

Caffeine is not classifiable as to its carcinogenicity to humans (Group 3).

For definition of the italicized terms, see Preamble Evaluation.


Last updated: 17 November 1997

    See Also:
       Toxicological Abbreviations
       Caffeine (ICSC)