487 diated transport mechanism, in which the liver plays a part: the decreased glucose uptake might be due to sub- clinical liver dysfunction undetectable by liver-function tests, Heavy metals may have a triple effect on the brain-by direct action on nervous tissue, by damaging the blood/brain barrier, increasing its permeability, and by damaging other organs, such as the liver. However, other influences are involved. For example, we cannot explain why patient 1 deteriorated so dramatically dur- ing pregnancy or why she improved on calcium and magnesium injections. We conclude that the dangers of cosmetics containing Bi and Hg used to lighten the skin should be better appreciated. Small amounts used for long periods can result in accumulation and intoxication, including ence- phalopathy. Patients with atypical clinical pictures, for whom the label "neurasthenic with psychogenic symp- toms" is tempting may have been exposed to a toxin, and history taking should include questions about con- tact with heavy metals. Atypical, unexpected, or para- doxical neurological reactions to a drug (as in our cases) should suggest a fault in its metabolism; this could be toxin induced. We thank Prof. F. Reimer and Dr R. M. Thiele for letting us study patients under their care and for their advice; and Miss B. Laatz for secretarial help. G. K. was supported by a vascular diseases research grant from the Thiemann Co. Requests for reprints should be addressed to G. K. REFERENCES 1. Levaditi, C. Presse méd. 1922, 59, 633. 2. Zieler, cited by von Sabro, A. in Handbuch der Neurologie (edited by Bumke and O. Foerster); p. 380. Berlin, 1935. 3.Beliles, R. P. in Toxicology (edited by L. J. Caserett and J. Doull); p. 467. New York, 1975. 4. Burns, R., Thomas, D. W., Barron, V. J. Br. med. J. 1974, i, 220. 5. Australian Drug Evaluation Committee, Med. J. Aust. 1974, ii, 664. 6. Buge, A., Rancurel, G., Poisson, M., Dechy, H. Nouv. Presse méd. 1974, 3, 2315. 7. Cambier, J., Masson, M., Dairou, R. ibid. 1974, 3, 2662. 8. Lhermitte, F., Degos, C.-F., Signoret, J.-L. ibid. 1975, 4, 419. 9. Raverdy, P., Smagghe, A., Marie, J.-P., Prinseau, J. ibid. 1975, 6, 1050. 10. Tournilhac, M., Flor, B., Dordain, G., Roye, J. M. ibid. 1975, 4, 430. 11. Wüstner, H., Orfanos, C. E. Dt. med. Wschr. 1975, 100, 1694. 12 Summa, J.-D. Münch. med. Wschr. 1975, 117, 1121. 13 Alexander, A., Mendel, K. Dt. med. Wschr. 1923, 31, 1021. 14. Hoyer, S., Oesterreich, K., Weinhardt, F., Krüger, G. J. Neurol. 1975, 210, 227. 15. Kety, S. S., Schmidt, C. F.J. clin. Invest. 1948, 27, 476. 16. Hoyer, S. Klin. Wschr. 1970, 48, 1239. 17. Bourdon, R., Galliot, M., Prouillet, F. Ann. Biol. clin. 1974, 32, 413. 18. Urizar, R., Vernier, R. L.J. Am. med. Ass. 1966, 198, 207. 19. Quadbeck, G., Hoyer, S. in Research on the Cerebral Circulation (edited by J. S. Meyer, M. Reivich, H. Lechner, and O. Eichhorn); p. 196. Spring- field, Illinois, 1972. "In this symposium, people have spoken of the preventive art as xmg something separate from the therapeutic art. I find it difficult ’3 believe that the subject of preventive medicine can flourish in the absence of an understanding of disease and an understanding of dis- ease means the use of scientific methods... A rigorously scientific study of the effects of smoking and raised arterial pressure was first needed before we could even talk about prevention in these areas. In :’. view, the job of academic clinical researchers is to understand and teach in relation to preventive medicine. Implementation of many :f our deductions may not even be a job for doctors. If the superbly ramed health visitors we have been hearing about would accept a new :; then I am sure they would have a much greater impact in preven- medicine than could be achieved by any group of doctors... Why it be thought that university clinical departments need to focus their attention on implementation of preventive medicine?"-A. Guz ;^. Health Care in a Changing Setting (Ciba Fndn Symp. no. 43); p. :". Amsterdam, 1976. MULTIPLE IMMUNE COMPLEXES AND HYPOCOMPLEMENTÆMIA IN DERMATITIS HERPETIFORMIS AND CŒLIAC DISEASE IDRIS MOHAMMED E. J. HOLBOROW Medical Research Council Rheumatism Unit, Canadian Red Cross Memorial Hospital, Taplow, Berkshire LIONEL FRY BERNADETTE R. THOMPSON Department of Dermatology, St. Mary’s Hospital, London W2 A. V. HOFFBRAND Department of Hœmatology, Royal Free Hospital, London NW3 J. S. STEWART West Middlesex Hospital, Isleworth, Middlesex Summary Circulating immune complexes have been detected in 100% of 59 patients with dermatitis herpetiformis (D.H.), and in 100% of 27 patients with cœliac disease (C.D.). Three methods for detecting immune complexes were employed: radiobio- assay, which gave an incidence of 77% in D.H. and 81% in C.D.; C1q binding activity, with which the incidence was 83% and 96%, respectively; and precipitation with 4% polyethylene glycol (69% positivity in D.H., 100% in C.D.). The immune complexes in D.H. and C.D. were com- pared with those in sera from 23 patients with systemic lupus erythematosus (S.L.E.). Multiple complexes of dif- fering properties were found in D.H. and C.D. but not in S.L.E. The varying nature of the complexes in D.H. and C.D. may account for the damage to different tissues (skin, small intestine, reticuloendothelial system). Low third component of complement was found in 49% and low C4 in 20% of D.H. patients. C3 hypocomplementæ- mia was found in 26% of patients with C.D. Introduction SEVERAL immunological abnormalities have been de- scribed in dermatitis herpetiformis (D.H.) and coeliac dis- ease (C.D.).1 z In a previous study the anticomplementary method revealed circulating immune complexes in 12 of 15 patients with D.H.3 With three different tests we now show that immune complexes are present in the sera of all patients with D.H. and C.D., and that these complexes are multiple. In addition, we record hypocomplementæ- mia-a finding not hitherto reported in these diseases. Patients and Methods 83 specimens of sera from 59 patients attending a dermati- tis-herpetiformis clinic were studied. 2 specimens were examined from each of 24 patients and 1 specimen each from 35 patients. Sera from 27 patients with coeliac disease who were completely controlled on gluten-free diet were also tested. In addition, sera from 60 normal volunteers of similar age and sex were studied. D.H. was diagnosed by clinical and histologi- cal features, response of the rash to dapsone, and the presence of IgA deposits in the uninvolved skin. The age of the patients ranged from 21 to 75 years and there were 36 males and 23 females. 14 sera were from patients whose rash was completely controlled by a gluten-free diet (G.F.D.), 28 from patients tak- ing a G.F.D. but who were also taking dapsone for control of their rash, and 41 from patients taking only dapsone. The