T H E LANCET
bacteriological cultures-are essential. Only in rare circumstances should a test be interpreted outside the clinical setting in which it is done. T h e test merely serves to intensify, confirm, or weaken our suspicion that the patient has the disease. The patient described by Maderazo had apical infiltrates on a chest radiograph. T h e number of diagnostic possibilities is limited, and excluding tuberculosis on the basis of an equivocal skin test was inappropriate. David J Evans, Richard J Barker, *Duncan M Geddes Royal Brompton Hospital, London SW3 6NP. U K
Maderazo EG. Interpreting tuberculosis skin tests. Lancer 1996;348: 832-33. 2 Holden M, Dubin MR, Diamond PH. Frequency of negative intermediate-strength tuberculin sensitivity in patients with active tuberculosis. N E n g l J M e d 1971;285: 1506-09. 3 Odelwo EO.Mantoux test tuberculin reaction in Nigerians. AfrJ Med Med Sci 1994;23: 91-98. 4 Rose DN, Schechter CB, Adler JJ. Interpretation of the tuberculin skin test. J Gen Intern Med 1995; 10: 63542.
with the then new non-sedative H, antagonists terfenedine, and astemizol had n o effect on the pruritus of eczema (measured subjectively as itch and objectively as scratch, with the use of limb meters), whereas old and less potent, but centrally depressive HI-antagonists such as trimeprazine were antipruritic. Furthermore, the central action of antipruritic antihistamines and other drugs is not solely sedative: nocturnal itch-provoked scratch was not suppressed by barbiturates (although they may be amnesic) but was suppressed by anxiolytic benzodiazepines such as nitrazipam and certain other central depressants.’ As Greaves and Wall say, much remains unknown about the mechanism of itch and its control (and certainly, greater use of objective methods of measurement would help’,’), which is why it would be sad if an important therapeutic lesson was lost: the pure HI-antihistamines have no effect on the itch of eczema. Sam Shuster Department of Dermatology, University of Newcastle upon Tyne. Newcastle upon Tyne NE2 ~ B w U, K Greaves MW, Wall I‘D. Pathophysiology of itching. Lancet 1996;348: 93840. 2 Kraus L,Shuster S . Mechanism of action of antipruritic drugs. BMJ 1983;287: 1199-200. 3 Shuster S. Reason and the rash. h c R Inst Great Britain 1981;53: 136-63. 1
Pathophysiology of itching .%-Greaves and Wall (Oct 5, p 938)’ suggest that “opioid p-receptor-dependent processes activate inhibitory circuits in the central nervous system and regulate the extent of intensity and quality of perceived itch”. Although it is true that stimulation of opioid preceptors inhibits pain, the evidence suggests quite the opposite for itch: activation of opioid-dependent pathways (for example, by morphine) has a stimulatory, rather than inhibitory, effect o n itch.” As would be predicted from this observation, molecules that block opioid receptors decrease itching in several clinical and experimental systems;’ this is the converse of what happens with pain, and indicative of the curious, and at least partly reciprocal, relation between the two sensations. Perhaps the same circuits are inhibitory for one and stimulatory for the other, but it is also possible that different circuits are involved. I admit that in the absence of neuroanatomical proof, thinking about pathways of inhibition and disinhibition interacting with agonists and antagonists is akin to working with double and triple negatives, open to multiple theoretical models, and enough to make one’s face,’ if not head, itch.
SIR-We question whether the p-system, as Greaves and Wall’ state, is the only relevant central opioid receptor involved in modulating itching. Our opinion is based on the findings in patients with primary biliary cirrhosis. In this condition there is a sustained rise in plasma concentration of Gopioid agonists (met-enkephalin and leu-encephalin).’ This increase allows these substances to penetrate the blood-brain barrier. Thomton and Losowski’ reported that following first administration of the opioid antagonist nalmefene itch was much alleviated, but the improvement was accompanied by a simultaneous opioid withdrawal reaction. They felt that the improvement in the level of pruritis was probably caused by the inhibition of the opioid substances that had caused the release of pruritic substances. It is of course possible that these centrally active Gagonists contributed to itch by direct action on the central nervous system &receptors, especially since the therapeutic effect of the antagonist occurred with the initial dose.
Jeffrey D Bernhard
*Mark A Gillman, Frederick J Lichtigfeld
Division of Dermatology, University of Massachusetts Medical School, Worcester,
South African Brain Research Institute. Waverley 2090, Johannesburg, South Africa
MA 01655. USA
Greaves MW, Wall PD. Pathophysiology of itching. Lancet 1996;348: 938-40. 2 Thomton JR, Losowski MS. Opioid peptides and primary biliary cirrhosis. BMJ 1988;297: 1501.03. 1
Greaves MW, Wall PD. Pathophysiology of itching. Lancet 1996;348: 93840. Tuckett RP.Neurophysiology and neuroanatomy of pruritus. In: Bemhard JD, ed. Itch: mechanisms and management of pruritus. New York McGraw-Hill, 1994: 1-22. Lemer EA.Chemical mediators of itching. In: Bemhard JD,ed. Itch: mechanisms and management of pruritus. New York: McGraw-Hill, 1994:23-35. Bergasa NV,Jones EA. The pruritus of cholestasis: potential pathogenic and therapeutic implications of opioids. Gastroenterology 1995;108: 1582-88. Thomas DA, Williams GM, Iwata K, Kenshalo DR, Dubner R. Multiple effects of morphine on facial scratching in monkeys. Anesth Analg 1993;17: 933-35.
SIR-In an otherwise useful review, Greaves and Wall’ say “the extent to which [the antipruritic effect of antihistamines in atopic eczema] is attributable to the side-effect of central sedation rather than local histamine antagonism in the skin is unclear”. But the reference cited’ was absolutely clear. It reported that total, or virtually total, HI-blockade in the skin
Vol348 * November 30, 1996
SIR-I believe that Greaves and Wall1 misuse the term tachykinin. They state that “Tachykinins include the neuropeptides substance P, calcitonin gene-related peptide (CGRP), and vasoactive intestinal peptide (VIP)”. Erspamer’ introduced the word tachykinin to describe the promptness of action on smooth muscle (Greek takus=fast) of peptides he isolated from the skin of several species of amphibians and the salivary glands of Eledone, a Mediterranean octopod. Among the peptides he included as members of this family are physalaemin, eledoisin, and kassinin. Substance P is a mammalian tachykinin, as are neurokinin A (substance K) and neurokinin B (neuromedin K). The structural feature usually regarded as a characteristic of this peptide group is a C-terminal sequence of the type Phe-X-Gly-Leu (or Met)-Met-NH,. T h e aminoacid at
position 4 from the C-terminus (X) is important for specificity. By contrast, C G R P is a member of the calcitonin/amylin family, and VIP of the PHI/GIP/glucagon/GLP-Usecretinfamily. None of these peptides contain the characteristic tachykinin C-terminal sequence. Both CGRP and VIP are vasoactive, which may contribute to their ability to potentiate substance P-induced itching. Christopher H S Mclntosh Department of Physiology. Faculty of Medicine. University of British Columbia, Vancouver BC. Canada V6T 123 1
Greaves MW, Wall PD. Pathophysiology of itching. Lancer 1996; 348: 93840. Erspamer V. Biogenic amines and active polypeptides of the amphibian skin. Ann Rev Fhannacoll971; 11: 327-50.
Inhibition of LDL oxidation by cocoa SIR-waterhouse and colleagues (Sept 21, p 834)’ report that chocolate contains substances that may be beneficial for arteries, apparently because of the abundance of flavonoid polyphenol. They found that 41 g of milk chocolate contains almost as much phenol as 140 m L of red wine.* It is known that cocoa is rich in polyphenols such as (-)epicatechin’ and that a coloured component of cocoa (cocoa-red) is polyphenol, which is found also in red wine. We have reported‘ that the phenol in red wine inhibits the oxidation of low-density lipoproteins (LDLs) and may help to prevent atherosclerosis, which is thought t o be provoked by oxidised LDL. We have also found that cocoa inhibits L D L oxidation in vitro. We measured the oxidation of L D L in a 1 m L reaction system consisting of 10-50 ng cocoa, 200 nmoL V70 (2,2’azobis[4-methoxy-2,4-dimethyl-valeronitrile]),and 68.6 pg LDL. Cocoa prolonged the oxidation lag time in a concentration-dependent manner. If cocoa is an active invivo antioxidant, then a diet that includes chocolate may also afford some protection against atherosclerosis. We therefore requested 12 male volunteers (mean age 39.0 [SE4.01 years) to consume 35 g of delipidated cocoa. Their venous blood was sampled before, and then at 2 h and 4 h after consuming the cocoa. LDL was isolated from each 1.1 m L plasma fraction by ultracentrifugation (density= 1.006-1.063 g/mL) and oxidation of this LDL was followed by measuring the conjugated dienes formed with V70. T h e L D L oxidation lag time before ingesting 35 g of cocoa was 61.2 (SE 6.4) min, but was prolonged at 2 h after cocoa (70.3 [6.1] min); at 4 h after cocoa, the initial lag time was similar to that before consumption (64.2 [5.6] min) (figure). Our results indicate that cocoa inhibits LDL oxidation 2 h after consumption. Thus we have shown that chocolate inhibits LDL oxidation in vivo. Hertog and colleagues’ 8o
I * i
Time after administration (h) Figure: Changes of plasma LDL oxidation lag time after 35 g of delipidated cocoa n=12.Points are means (SE). *p<0.005 by paired t-test.
reported that intake of flavonoid-rich foods (eg, tea, onion, and apple) may also protect against coronary heart disease. The results of Waterhouse and colleagues,l supported by our data, suggest that dietary intake of cocoa rich in antioxidant flavonoids may reduce the risk of atherosclerosis, and morbidity and mortality from coronary heart disease. T h e caffeine content of cocoa is 0.009% by weight, compared with coffee (0-04%), black tea (0.06%), and green tea (0.01 %). Theobromine, another natural product in cocoa which can influence haemodynamics, has an LD,, in rats of 837 mg, which is equivalent to 200 50 g bars of milk chocolate, 90 of black chocolate, and 70 cups of hot chocolate ingested by a 60 kg human. To avoid any sideeffects of cocoa, excessive chocolate consumption is not recommended. A similar case for moderation can be made for red wine, so as to avoid liver cirrhosis or hypertension: too much is as bad as too little. We thank the Uehara Memorial Foundation, Japan Health Sciences Foundation and Aging and Health, and the Ministry of Health and Welfare of Japan for funding part of this work.
* K Kondo, R Hifano, A Matsumoto, 0 lgarashr, H ltahura *Division of Clinical Nutrition, National Institute of Health and Nutrition and Insitute of Environmental Science for Human Life. Ochanomizu University, Tokyo 162.Japan
Waterhouse AL, Shirley JR, Donovan JL. Antioxidants in chocolate. Lancet 1996; 348: 834. Frankel EN,Waterhouse AL, Teissedre PL. Principal phenolic phytochemicals in selected California wines and their antioxidant activity in inhibiting oxidation of human low-density lipoproteins. J Ag Food Chem 1995; 43: 890-94. Kim H,Keeney PG. (-)-Epicatechin content in fermented and unfermented cocoa beans. JFood Sci 1984; 49: 1990-92. Kondo K, Matsumoto A, Kurata H, er al. Inhibition of oxidation of low-density lipoprotein with red wine. Lancet 1994; 344: 1152. HeKog MGL, Feskens EJM, Hollman PCH, Katan MB, Kromhout D. Dietary antioxidant flavonoids and risk of coronary heart disease: the Zutphen Elderly Study. Lancet 1993; 342: 1007-1 1.
Sleep disturbance and blindness SIR--An accumulating body of evidence-contributions to which have featured prominently in The Lancetl,’-have led to a general consensus among sleep researchers that visual impairment may predispose individuals to sleep disturbance. This conclusion has an intuitive appeal and perhaps this is where alarm bells should start to sound. Nonetheless, prevalence studies have consistently found sleep quality in the visually impaired to be significantly poorer than in the normally sighted population,’’ and this finding is now underpinned by an explicit chronobiological hypothesis: that the light-dark cycle is the most powerful entrainment mechanism of circadian rhythms, including that of sleepwake. Our concern is that sleep researchers, perhaps discerning that a single brutal fact can slay a beautiful theory, seem to have studiously avoided considering alternative hypotheses that might account, wholly or in part, for the sleep problems of the blind. We have analysed the questionnaire responses of 1 139 guide-dog owners in the UK from a survey in which we inquired into their ophthalmic and systemic health, including sleep quality, and their lifestyle (eg, alcohol and tobacco consumption, exercise). 158 (14%) reported themselves to be depressed, and of these 74 (47%) described their sleep quality as either poor or very poor. Of the 981 not depressed, only 134 (14%) reported sleep disturbance. In a multivariate logistic regression, depression was among the most significant predictors (p
Vol348 November 30. 1996