212 Letters to the Editor COLD-RELATED DEATHS IN SOME DEVELOPED COUNTRIES SIR,-During the cold spell in Britain last February newspapers reported that the death rate was 1000 per week above normal. Such statistics are not recorded for other countries, even those with much colder winters, such as Canada and Sweden. Few people die of cold as such, but many die of related illnesses. Cold-related deaths amount to 12% of total deaths in England and Wales, 11 % in Scotland, but only 4-6 % in the other countries investigated (table). COLD-RELATED DEATHS IN SOME DEVELOPED COUNTRIES Source: Umted Nations DemographIc Yearbook (1980), table 30. Total deaths in year less 12 times the monthly deaths for July to September, expressed as number of deaths (and as % of total). Since about 500 people per million of population die of cold each year and since the population of Europe, the USSR, and North America is about 1000 million, it can be estimated that about half a million people in these regions die of cold in a year. It seems that cold is to the developed countries what malnutrition is to the developing countries. Neither kills directly-they kill by proxy. Cold-related deaths as % of total are to some extent poverty indicators but I doubt if the high figure for Britain is entirely due to poverty. Even wealthy people in that country shiver their way through winter, and I am not merely referring to unfortunates in unheatable stately homes. Cold-related death is serious numerically in developed countries and in Britain in particular; its prevention lies not only in poverty alleviation but also in education. Apt 814E, 4201 Cathedral Avenue NW, Washington DC 20016, USA MIKAEL GRUT ACYCLOVIR IN ACCELERATED PHASE OF CHEDIAK-HIGASHI SYNDROME SiR,—Chediak-Higashi syndrome (CHS) is an autosomal recessive disorder characterised by giant lysosomal granules, partial albinism, and recurrent infections. An accelerated, lymphoproliferative phase of this disease, with fever, massive hepatosplenomegaly, coagulopathy, and pancytopenia, heralds death in most patients.1,2 The accelerated phase has been associated with seroconversion and an abnormal antibody reponse to Epstein- Barr virus (EBV).3,4 Antibodies to the viral capsid antigen (VCA) reach high titres that are often matched by similarly high levels of antibodies to the restricted component of the early antigen (EA-R).3,4 High antibody titres to EA-R suggest a persistent, active EBV infection with the release of infectious EBV particles. Because acyclovir has had some benefit in the acute, lytic phase of EBV infection,’’ a child in the accelerated phase of CHS was treated with this drug. At age 2! years this child had varicella complicated by encephalitis. At age 3 (on Nov 4, 1985) he was admitted with fever, hepatosplenomegaly, coagulopathy, pancytopenia, and an EBV- specific antibody profile characteristic of the accelerated phase (table) of CHS. The patient was treated with prednisone 2 mg/kg daily and a 10-day course of intravenous acyclovir 500 mg/m2 three times daily. The fever and coagulopathy resolved and the SERIAL ANTIBODY TITRES AND IMMUNOGLOBULIN LEVELS haemoglobin concentration and platelet count were increasing when he was discharged, on tapering doses of prednisone and oral acyclovir (400 mg/m2 thrice daily). The patient was doing well when seen on March 13,1986, but he was readmitted 6 weeks later with a relapse of the accelerated phase. The patient was treated with intravenous prednisone, acyclovir, and etoposide. His hospital course was complicated by longlasting pancytopenia, staphylococcal sepsis, and gastrointestinal haemorrhages, but he slowly improved and was discharged on July 3 on oral acyclovir and prednisone. On Sept 17 the patient was readmitted for chemotherapy in preparation for a bone marrow transplant but he died of disseminated aspergillus infection. He had IgG and IgA antibodies to EA-R and, less strikingly, to VCA; titres fell in parallel with clinical improvement and rose during relapse. The terminally low titres reflect, at least in part, advancing hypogammaglobulinaemia. These data provide further evidence that EBV plays an important role in the emergence of the accelerated phase of CHS and suggest that acyclovir has some beneficial effect in this disorder. Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania 19104, USA; and Joseph Stokes, Jr Research Institute, University of Pennsylvania School of Medicine MARY ELLEN CONLEY WERNER HENLE 1. Blume RS, Wolff SM. The Chediak-Higashi syndrome. Studies in four patients and a review of the literature. Medicine 1972; 51: 247-80. 2. Rubin CM, Burke BA, McKenna RW, et al. The accelerated phase of Chediak- Higashi syndrome. An expression of the virus-associated hemophagocytic syndrome? Cancer 1985; 56: 524-30. 3. Merino F, Klein GO, Henle W, Ramirez-Duque P, Forsgren M, Amesty C. Elevated antibody titers to Epstein-Barr virus and low natural killer cell activity m patients with Chediak-Higashi syndrome Clin Immunol Immunopathol 1983; 27: 326-39 4. Merino F, Henle W, Ramirez-Duque P. Chronic active Epstein-Barr virus infection in patients with Chediak-Higashi syndrome J Clin Immunol 1986; 6: 299-305 5. Sullivan JL, Medveczky P, Forman SJ, Vaker SM, Monrow NS, Mulder C. Epstein-Barr virus induced lymphoproliferation: Implications for antiviral chemotherapy. N Engl J Med 1984; 311: 1163-67. 6 Andersson J, Britton S, Ernberg I, et al. Effect of acyclovir on infectious mononucleosis. a double-blind, placebo-controlled study. J Infect Dis 1986, 153: 283-89 PROSTAGLANDINS, SMOKING, AND DUODENAL ULCERS SIR,-Dr Hawkey and Dr Walt (Nov 8, p 1084) state that our reportl that the ill-effects of cigarette smoking on duodenal ulcer healing were nullified by E-type prostaglandins could be a chance finding, and they cite studies2-6 that they think are negative in this respect. This seems a sweeping statement that could bury an important scientific advance. One of the studies cited by Hawkey and Walt as evidence that prostaglandins do not help to overcome the adverse effects of smoking on duodenal ulcer healing is that by Brand et al.2 These workers showed that, among the 100 patients on placebo, 44% of the 63 smokers and 65% of the 37 non-smokers had their duodenal ulcer healed in 4 weeks (p = 0-038), whereas among the 107 patients treated with misoprostol 200 pg four times daily for 4 weeks, 73 % of 60 smokers and 79 % of the non-smokers had their ulcers healed. These findings confirmed ours:1 in a 12-week study with repeat