Absinthism: fictitious 19th century syndrome
Page 12 of 15
Toxicological rehabilitation of absinthe
Until recently, the thujone content of pre-ban absinthe was largely unknown and was calculated in 1992 by Arnold [ 20 ] to be as high as 260 mg/l (a value very often cited in the newer literature, e.g. Refs. [ 10 , 16, 46 , 56 , 87, 88 ]). The value of 260 mg/l was determined on the basis that 100 l of absinthe employed 2.5 kg of dried Artemisia absinthium (1.5% oil, of which 67% is thujone; corresponding to 251 mg/l of thujone in the final product) and 1 kg of dried Artemisia pontica for col-oration (0.34% oil, of which 25% is thujone, correspond ing to 9 mg/l of thujone in the final product) . Max independently calculated a similar concentration [83 ].These calculations assumed that the total amount of thujone would be recovered in the final product. The following three points were given by Arnold to support his calculation of relatively high concentrations. First, by adding water to the first decoction before heating, a type of "steam-distillation" was achieved wherein the amount of any constituent distilled over depends on both its vapor pressure and molecular weight. In this way the effect of a low vapor pressure for a particular compound may be counteracted to some extent by its high molecular weight relative to that of water. Second, the distillation head of the industrial apparatus was simple and little attempt was made to restrict carry-over by aerosol entrainment. And third, the purpose of the secondary extraction at moderate temperature was twofold, to achieve a green coloration and to add additional flavor [ 20 ].
However, it cannot be totally disregarded that during distillation a discrimination of thujone occurs. Historic recipe books prescribed the removal of the heads and tailings [ 26 -28 ]. Duplais, for example, indicates that after maceration in 16 l of alcohol (85%vol) and addition of 15 l of water, only 15 l of product are withdrawn. 1 l of alcohol is discarded in the process . In a non peer-reviewed magazine article, Turner described first experiments conducted by T. Breaux on a French distillery built in 1834. After distillation in a historic still built for absinthe, the thujone originally present in the macerate was not recov ered in the distillate [71 ]. The thujone content of absinthe can then only be caused by the second coloration step, which would lead to a concentration of 9 mg/l according to Arnold's calculation.
Baker reports another calculation that resulted in thujone concentrations of 60â90 mg/l . Wilson  estimated in 1936 that absinthe made from essences contained 1.8 to 45 mg/l, and absinthe made with wormwood contained 2 to 34 mg/l of thujone.
Hutton pointed out that the thujone content of pre-ban absinthes could have been overestimated because of the insufficient analytical methods that were available at the time [ 88]. Historically applied methods for the determi nation of levels of thujone in absinthe were based upon iodometric titration  or color reactions ; these sometimes provided only detection limits as high as 20 mg/l and were therefore unfit for the detection of small quantities. At the beginning of the 19th century, the most modern methods were based upon the reaction of thujone with sodium nitroprusside, sodium hydroxide and acetic acid and provided a limit of detection of 5 mg/l [ 92 ]. However, this color reaction was highly unspecific and therefore other essential oils, aldehydes and ketones led to a similar reaction to thujone. Even by improved sample preparation, it was not possible to avoid these interferences. A positive reaction in the case of thujone analysis could not automatically be interpreted in such a way as to prove that the spirit in question was made with wormwood. However, a negative result was regarded as proof of the absence of wormwood oil .
The sensitive and selective determination of thujone in spirits was only possible by using modern chromatographic methods. The first gas chromatographic method with a flame ionization detector for the determination of thujone in alcoholic beverages was developed by Mérat et al. in 1976 . In a recent study of our working group
[ 95 ] absinthes produced according to historic recipes did only contain relatively low concentrations of thujone (mean: 1.3 Â± 1.6 mg/l, range: 0â4.3 mg/l), which is 50â 100 times below the NOEL (No observed effects level) of thujone for convulsions determined in animal experiments. A vintage absinthe from Tarragona (1930) showed a relatively low thujone concentration of 1.3 mg/l. Swiss Val-de-Travers absinthes from traditional small distilleries contained 9.4 and 1.7 mg/l of thujone. Krumm et al. verified our results by their production of absinthes after authentic French recipes. All manufactured products had thujone concentrations below 1.5 mg/l . Hutton found 6 mg/l of thujone in a Pernod absinthe from 1900 [ 88 ]. In a non peer-reviewed magazine article, Ashcraft reports tests on pre-ban absinthes conducted by T. Breaux, who found thujone amounts around 5 mg/l . Schaefer et al. found such low thujone concentrations (<0.01 mg/ l) in a legal French absinthe dating from 1904 that the authors even proposed the "toxicological rehabilitation" of absinthe [98 ]. In a current study of the neuropsychiatric toxicity of absinthe by Luauté et al. it was concluded that recent toxicological studies do not prove, any more than in Magnan's time, that the beverage itself was epilep togenic . The toxicity of pre-ban absinthes, as that of modern ones, was found to be essentially due to their alcohol content.
The theory of a gross overestimation of the thujone content was, therefore, verified by six independent studies [ 88 , 95- 99]. The discrepancy between the experimental findings of pre-ban absinthes (low thujone concentrations) and the calculations of Arnold and Max (high thu jone concentrations) [20,83] could not be resolved so far. Further research is needed to study the behavior of thujone during distillation. Considerable differences to the composition of the distillate may result between batch-wise distillation of diluted alcohol and fractional distillation of an undiluted macerate.
Currently no experimental evidence does suggest that historic absinthes had such high thujone contents to cause toxic effects. On the contrary, the analyzed historic products appear to have complied with today's maximum limits derived to exclude toxic or other unwanted effects. The feared return of absinthism, proclaimed e.g. by Hein et al. [ 100 ], Holstege et al.  or MÃ¼ller  is therefore exaggerated. The effects of the recent types of absinthe are predominantly caused by the naturally high alcoholic strength (>50%vol), although it is possible to reach effective thujone blood levels, if illegally produced and distributed absinthe is ingested.