In an article in Food research International, Vol 29, 325-330 (1996), the content of caffeine in various teas and styles of brew were compared.
One of the formosa oolongs had less caffeine than the green and black teas on a dry weight basis, but when all were prepared to supplied directions, they had similar caffeine levels. When brewed for a series of three repeated brews, the caffeine concentration was higher in the first brew for one of the other Formosa oolongs than other green, black, and oolong teas. The caffeine content decreased dramatically from the first to the third brew, about 50-70mg/cup in first cup, 15-25 in second, and 5-10 in third, each brew about 5 minutes in length.
Bag teas had slightly less caffeine than loose teas. This could be due to the use of broken and poorer quality leaves in bags.
Theobromine was also measured, which was very low: 1.5-3mg/cup for first brew, 0.5-1 in second, 0.2-0.5 in third. Black teas had slightly more theobromine, attributed to the fact that younger leaves have more theobromine, and the young leaves are chosen for black tea production due to higher polyphenol levels.
One other interesting result was that no (less than detection limit, 0.1mg/cup) theophylline was detected in any of the brews.
This paper challenges many of the popular assertions about the caffeine content of teas, in that:
1. All teas have roughly similar caffeine contents, and one cannot rely on the belief that green tea has less caffeine, as asserted by many popular claims. One such claim is that degree of fermentation governs caffeine content, with green having 1/3 and oolong having 2/3 the caffeine content of black tea. This has been debunked repeatedly. [See new information at the end of this section] Small differences in average caffeine contents of green vs black tea may be due to the use of China varietals for green and Assam varietals for black, a line which comes closer to elimination with Indian green teas and oolongs. Another is the use of younger or more mature leaves, where there is some variation. Caffeine is a product of the natural metabolism of RNA in the plant and thus will vary with senescence. Some caffeine may also be made during fermentation, but it is not a sufficiently orderly process to clearly delineate such even multiples as 1/3, 2/3, etc.[See new information at the end of this section] Fermentation time is decided by taste and plantation experience, not by biochemistry, and is not controlled to scientific tolerances.
2. Caffeine content does depend on brew technique and leaf size and variety, and that without an accurate analysis as such available from a qualified laboratory, one cannot assert that some teas have a lower or greater caffeine dose.
3. One popular assertion suggests that tea can be decaffeinated by brewing quickly and discarding the first brew. Since tea must be decaffeinated greater than 90% to be considered decaf (and most commercial decafs are >98% decaffeinated), the second brew, although lower in caffeine, is *not* decaffeinated (65-75% of caffeine coming out in the first five minute brew).
Making informed decisions about what is a suitable level of caffeine in one's diet is a personal decision, and although doses of less than 200mg/day of caffeine have generally not been associated with health risks, one should *always* consult with a doctor and *not* consider any of this (or other literature) to be advice on your personal health.
I urge all tea drinkers to be wary of advice on caffeine levels in tea, particularly results which say that green or oolong tea has less caffeine.
(February 11, 2000)
Caffeine varies in the fresh green leaf depending on fineness of pluck. The fourth leaf (Coarse) has about one third less than the first leaf (Very Fine Pluck). Caffeine varies with type -seedling tea (3-5%) has perhaps twice that of clonal tea (1-3%). As most green tea is seedling (i.e bushes derived from seeds) and much black tea is clonal (bushes derived from VP - vegetative propagation) this means that black teas have often, weight for weight, less caffeine. But also green teas are often liquored with less leaf.
Caffeine is also variable depending on season - some clones hit 5% in December, down to 2% in July, in Kenya. And on nitrogen fertiliser which can add a tenth to the normal caffeine level.
Incidentally caffeine level can be manipulated in the factory - withering temperature peak for caffeine is 25 deg C, ten degrees above and below this reduce caffeine by 0.2 to 0.3% ponts. Likewise withering time - longer wither gives more caffeine, straightline response up to 30 hours (not that anyone would normally wither more than 20 hours). However, length of ferment decreases caffeine, down from 3.2% at zero minutes to 2.8% after 90 minutes,