Document 0aRrBXZEZY5o667ZqQGgNdEn

<Jo: Dr. C. C. Pfeiffer (2^onatcl cR. -Cyncun, /"& (2^>. N 31002 an jvruminju *w7 V UUUACK 20'* ) ,^/Blood lead concentration, blood pressure, and renal function ;. &. I ! i S J POCOCK, A G SHAPER, D ASHBY, T DELVES, T P WHITEHEAD Abstract Blood lead concentrations were related to blood pressure and indicators of renal function in a clinical survey of 7735 middle aged men from 24 British towns. There was no overall evidence that blood lead concentrations were associated with systolic or diastolic blood pressure (r= 4-0-03 and 4-0-01, respectively). In the 74 men with a blood lead concentration of 1-8 ftrnol/l (37-3 frg/100 ml) or more there was some suggestion of Increased hyper tension,- but this did not reach significance. Blood lead concentration did not have any relation with serum creatinine concentration. Moderate increases in blood lead concentration were associated with small increases In mean scrum urate concentration and small decreases In mean serum urea concentration; these associations were both reduced when alcohol consumption was taken Into account. There is no indication that exposure to lead at con centrations commonly encountered in British men is responsible for impaired renal function or Increased blood pressure. Introduction * c *u:-r" - j Considerable public concern has been shown abo^p-the effects on health of exposure to lead both in adult^atiSTin children. In response to reports.that moderatejnefeases in blood lead concentration are associated witfrjrypertension and impaired renal function*1-' we present Ajtdmgs from a clinical survey of British middle aged mea^pdating measurements of blood lead concentration to blood pressure and serum creatinine, urate, and urea concentrations. Patients and methods ' J?;' "" '- The Regional Heart Study recruited 7735 men aged 40 to 59 who were randomly selected from representative general practices in 24 British towns. Details of the selection of towns and general practices and the methods of screening and data collection have been reported.* Each man's blood pressure was measured twice in succession with the London School of Hygiene sphygmomanometer, with the subject seated and his arm supported on a cushion. Diastolic blood pressure was recorded at disappearance of sounds (phase V). The mean of two readings of blood pressure was adjusted for observer variation within each town to allow for any inconsistencies among the three observers. -` Blood samples for biochemical analysis and haematological studies were taken into evacuated tubes. All samples reached the Wolfson Research Laboratories, Birmingham, by the next morning, and estimations were completed by noon. Serum concentrations of creati nine, urate, and urea were measured ou a Technicon SMA 12/60 analyser. Blood lead concentrations were analysed at the University of Southampton with frame microsampling atomic absorption spectroscopy.* A stria quality control protocol was maintained and the performance of the laboratory was continually monitored by participation in national and international quality assessment schema for analysis of blood lead concentration.. . Alcohol consumption Was recorded using'questions on frequency, quantity, and type similar to those in the General Household Survey.* A drink was defined as half a pint of beer, one glass of wine, or a single tot of spirits. For data analysis, eight drinking categories were used;, non-drinkets, occasional drinkers, weekend drinkers (one or two, three to six, or more than six drinks a day) and duly drinkers (one or two, three to six, or more than six drinks a day). Results ' " BLOOD LEAD CONCENTRATION AND BLOOD PRESSURE - - Blood pressure and blood lead concentrations were both measured in 7371 men (95% of men screened). The correlation coefficients of blood lead with systolic and diastolic blood pressure were r= 4-0-03 and r* 4-0-01 respectively. The association with systolic blood pressure, although close to zero correlation,, was significant at the 1% level. In large scale surveys, however, it is the magnitude of an association, not its significance, that is biologically important. Figure 1 shows the mean and 95% confidence limits for systolic blood pressure for men categorised according to blood lad con centrations. There was no indication of higher systolic blood pressure 152' P-' Department of Clinical Epidemiology and General Practice, Royal Free Hospital School of Medicine, London NW3 2PF S I POCOCK, ms c , ran, reader in medical statistics A G SHAPER, f f c p, ntCPATH, professor of clinical epidemiology D ASHBY, ms c , ran, research fellow in medical statistics Department of Chemical Pathology and Human Metabolism, University of Southampton, Southampton T DELVES, p h d , pare, honorary senior lecturer in chemical pathology Wolfson Research Laboratories, Queen Elizabeth Medical Centre, Birmingham T P WHITEHEAD, PHD, PSCPATH, professor of clinical chemistry Coirespondencc to: Dr S J Pocodi. ft Blood lead t/jmot/IH06 0-6-09 09-1-1 ' 12-V4 VS-17 >16 No of men 1984 Nblooo{%d )pwreitshssJyestoiic 4,8 3381 1423 385 125 675(20) 349(251 95(25) 22(17) 22(30) >160mmHg I f ig 1--Systolic blood pressure foe 7371 middle aged men categorised according to blood lead concentration, i Represents unadjusted mean and 95% confidence limits. Represents mean adjusted for age* town, body mass, alcohol consumption, social class, and observer. . Conversion: SI to tradiuoruzi unm---Lead: 2 funol/ls;20*7 pg/100 ml. TEH 0412552 DUP050453604 N 31002.01 THTR TS A S IN G LE PHOTOCOPY MADE BY THE NEW YORK ACADEMY OF M E D IC IN E FOR At. JOURNAL VOLUME 289 6 OCTOBER 1984 t ' ith higher blood lead concentrations. Age, town, body mass ^ ^alcohol consumption, social class, and observer have all i .vfotisly been shown to be associated with systolic blood pressure.4 jrigurc 1 also shows the mean systolic blood pressure by blood lead ^oeentratiou, after adjustment for these other factors using analysis ofcovariance. Again, there was no evidence ofany association between blood lead concentration and systolic blood pressure. To examine the prevalence of hypertension in relation to blood lead concentration figure 1 also shows the proportion of men with systolic blood pressure over 160 mm Hg for each blood lead con centration. There was no significant trend in these proportions, though, of the small group of men with a blood lead concentration of 1-8 /imol/1 (37-3 fig/100 ml) or more, 22 (30%) had hypertension compared with 1559 (21%) of all the other men (p=0-08). Findings for diastolic hypertension (>100 mm Hg) followed a similar pattern. There was an increase in diastolic hypertension in the small group of men with a blood lead concentration of 1-8 fimol/2 (37-3 Mg/100 ml) or more--that is, 11 (15%) had hypertension compared with 648 (9%) of all the other men. This difference, however, was only mar ginally significant (p=0-07) and there was no evidence of a trend at blood lead concentrations below 1-8 ftmol/1 (37-3 pg/100 ml). 673 tively, the associations with urate and urea being highly significant (p<0-001). Once again, however, it should be noted that, despite the levels of significance, the magnitude of each correlation was small and unlikely to be of biological importance. Figure 2 shows means and 95% confidence intervals for serum urate and urea concentrations for men categorised according to blood lead concentration. Except for the small groups of men with a blood lead concentration of 1-8 jimol/1 (37-3 fig/100 ml) or more there were consistent trends with blood lead concentration, decreasing for urea and increasing for urate concentrations. Alcohol consumption has previously been shown to be associated with increasing serum urate concentration and decreasing serum urea concentration.' In this previous report, heavy drinkers--for example, those drinlung more than three pirns of beer daily--had a 13% increase in mean serum urate concentration and a 9 % decrease in mean serum urea concentra tion compared with occasional drinkers. Also, as previously reported,, heavy drinkers had a 30% increase in mean blood lead concentration.4 Thus alcohol consumption must be taken into account when associa tions between urea, urate, and blood lead concentrations are studied. Accordingly, figure 2 also shows the means for serum urate and urea concentrations adjusted for alcohol consumption using analysis of covariance. The trends with blood lead concentration are weakened but not eliminated after allowance for alcohol consumption. The partial correlation of blood lead concentration with serum urate concentration was reduced from +0-10 to +0.06. For serum urea concentration the correlation changed from --0-08 to --0-05 after allowing for alcohol consumption. ... ... : :! `----I-------- r------1-------- t----------- r 400-j 390- | 380- ' ' . <0 6 06-08 09-1-1 1-2-14 15-17 >18 Blood lead (AJmol/l) Fto 2--Serum urea and urate concentrations for 7364 men categorised according to blood lead concentration, w Represents unadjusted mean and 95% confidence limits. Represents mean adjusted for alcohol consumption. Conversion: St to traditional units--Lead: 1 pmol/)s20-7 fig/ 100 ml. Urea: 1 mmol/12:6 ug/100 ml. Urate: 1 umol/l 16-8 ugl 100 ml. BLOOD LEAD CONCENTRATION AND RENAL FUNCTION Three indicators of renal function (serum concentrations of creati nine, utate, and urea) were measured in 7364 men for whom data on blood lead concentrations were also available (95% of men screened). As creatinine and urea concentrations both had highly skew distri butions log transformations were used, in statistical analysis. The coueianon coefficients for blood lead concentration with creatinine, urate, and urea concentrations were 0-00, +0-10, and -0-08 respec D,i.sc. ussion. t* . . .\'. . BLOOD LEAD CONCENTRATION AND BLOOD PRESSURE i . In this large representative sample of British middle aged men there was no overall evidence that moderate increases in blood lead concentration were associated with increased blood pressure. This finding contradicts a previous case-control study done in Glasgow.1 In that study 135 subjects with hypertension were each paired with a normotensive subject matched for age and sex. In. a paired analysis of blood lead concentrations, significantly more of the men with hypertension had increased blood lead concentrations than did the men with normal blood pressure and there was a similar but non-significant trend in women with hypertension. These findings were the basis for a conclusion that high blood pressure is associated with high blood concentrations of lead and that-this might explain the high prevalence of cardiovascular disease in the west of Scotland. The possible bias in such case-control studies must, however, be considered. For instance, the important influences of alcohol consumption and cigarette smoking on blood lead concentration were not known at the time of the Glasgow study and thus were not taken into account. The higher blood lead concentrations -in some patients with hypertension may therefore have been due to their high intake of alcohol rather than any causal influence of lead on blood pressure. Another possible explanation is that blood lead concentrations were higher in the Glasgow study largely because of plumbosolvent water supplies, and any potential effect of raised blood lead concentration on blood pressure may have been present only above certain critical limits such as 1-8 pmol/1 (37-3 pg/100 ml). This would be. consistent with no association being found in a similar case-control study in Birmingham, is which blood lead concentrations tended to be lower.* There is a weak suggestion from the data of the Regional Heart Study that subjects with blood lead concentrations of 1-8 pmoJ/1 (37-3 pg/ 100 ml) or more may suffer more often from hypertension. As only 74 men (1%) had blood lead concentrations of 1-8 pmol/l (37-3 pg/100 ml) or more it seems reasonable to conclude that exposure to lead makes a negligible contribution to high blood pressure in the general male population of Great Britain. BLOOD LEAD CONCENTRATION AND RENAL FUNCTION The association between blood lead concentration and three indicators of renal function (serum concentrations of creatinine. TEH 0412553 i* DUP050453605 ! .T H IS I S A SINGLE PHOTOCOPY MADE BY THE NEW YORK ACADEMY OF M E D IC IN E FOR j I PURPOSE OF STUDY OR .RESEARCH I N L IE U OF LEN D IN G THE O R IG IN A L . and urea) was examined in 7354 middle aged British * men. After allowing for the influence of alcohol consumption on serum biochemistry there was no substantial association of eincreased blood lead concentration with renal function. There was a weak positive trend towards increased serum urate concentration at high blood lead concentration and a reverse . trend for serum urea concentration. These findings differ from those of a previous study carried out in Scotland on people selected because their homes were thought to have lead pipes or storage tanks.* In that study raised blood lead concentration was associated with raised serum urea and urate concentrations 1 and excessive lead in domestic water supplies was considered to have a harmful effect on the community's health.. A review paper then questioned their use of arbitrary cut oft points for serum urate and urea concentrations, particularly in an elderly population.* In addition this case-control study took no account of alcohol consumption, which is now known to have a. considerable effect on. both blood lead concentration r and indicators of renal function. Several studies have reported that occupational lead poisoning I can result in severe renal impairment. For instance, an investi gation of renal function was carried out in 102 Rumanian patients with lead poisoning attributed to occupational exposure who had a mean blood lead concentration of 4 pmol/1 (83 pgf 100 ml).1* Seventeen patients had chronic nephropathy, mostly after more than 10 years of exposure to lead. Thirteen of these patients had also developed hypertension. Studies such as this have stimulated inquiry into whether more moder ate exposure to lead in the general population could affect renal function or blood pressure. It has been suggested that, in areas where occupational lead nephropathy is common, patients seen at the stage of hyper tension or chronic renal impairment typically have normal blood lead concentrations; their lead overload is shown only BRITISH MEDICAL JOURNAL VOLUME 289 6 OCTOBER 1984 by bone biopsy or an cdetic acid excretion test. Thus the possibility that exposure to lead earlier in life might account for subsequent hypertension or renal disease cannot be excluded. Our cross sectional study, however, using current blood lead concentrations, has not shown any evidence that recent ex posure to lead at concentrations commonly experienced by British men is responsible for increases in blood pressure 01 impaired renal function. The British Regional Heart Study la supported by a programme, grant from the Medical Research CoundL Biochemical estimations were carried out at the Wolfson Research Laboratories under a programme supported by the Department of Health and Social Security. Blood lead concentrations were estimated in the department of chemical pathology and human metabolism. University of Southampton (Professor B Clayton), supported by the DHSS. Reference* > . , 1 Beeren DG, Enkino E, Robertson Lancet 1976ai:l-3. Blood lead snd hypetteufaa. 2 Beeven DO. Gruicfahaok JK. Yeoman WB, Carter OF, Celdber* A, Moore MR. Blood lead aad cadmium fo human feypezteoaion.1 Environ Toxicol 1980;4:251-60. Kegioiui Heart Scud?: cardiovascular nit factor* n xuxddle-szcd meet a 24 town*. J9r Med7 19B1;283:17SMJ6. 5 Delves HT. A microsampling method for rapid detenolsadoo of lead la Mood by fUme atomic absorption spectrophotometry* AnaJyit 1970;95:431-8. 6 Cummin* RO, Shaper AG, walker M, Wale CJ. Srooldnf tad drintto* by middle-aged British meat effect* ofsocial class snd town ofresidence. Sr Mod y 1981^:1497-502.. iG, rocock Sf. Walker M* el fll. Effects ef alcohol and tnokuii on blood lead in middl-a*f British men. Br MtdJ 1982^4^99-302. 9 Shaper AG. Cardiovascular disease sad trace Free R Soc Land (BwQ 1979;205:135-43. 10 LQis R* Grrrilescu N. Nestortscu 8, Domitrivi C, Roveata A. Nephropathy fa chronic lead poiaooiag. BrJ ind Med 1968^*5:196-202. {Acc*ps*U8fufyJ$S4) I Cardiac arrhythmias during rewarming of patients with accidental hypothermia ANDREW C RANKIN, ALAN P RAE Abstract ' ` Accidental hypothermia has a Ugh mortality and Is associated with 'cardiac arrhythmia*. To determine the 1 incidence of arrhythmias and their importance 22 patients with accidental hypothermia (core temperature <35"C) were studied by 12 lead electrocardiography and continuous recording of cardiac rhythm. Although 14 of the patients died (64%), only six died while hypo- ' thermic. Prolongation of rite Q-T interval and the pre sence of J waves were related to the severity of the j hypothermia. Supraventricular arrhythmias, including I atrial fibrillation, were common (nine cases) and benign. I Ventricular extrasystoles were also common (10 cases), " but ventricular tachycardia or fibrillation did not occur University Department of Medical Cardiology, Royal Infirmary, Glasgow G3I 2ER ANDREW C RANKIN, mk c p, registrar ALAN P RAS, mk c p, senior registrar Corrapoadeace to: Dr A Rinkin. during rewarming. In eight patients who died while being monitored the terminal rhythm was asystole. There was no correlation between the severity of hypo thermia or the rate of rewarming and the clinical outcome. In the absence of malignant arrhythmias there is no indication for using prophylactic aadarrhythmlc treat ment in patients with accidental hypothermia. The presence or absence of severe underlying disease is the main determinant of prognosis. ......... Introduction Accidental hypothermia is a common reason, for admission to hospital during the winter1 * and has a high mortality, especially in the elderly.* * The condition is associated with life threatening arrhythmias, in particular ventricular fibrillation, and there have been several reports of prolonged but successful resuscita tion in such patients.1-* Our aim was therefore to record continuously the cardiac rhythm of hypothermic patients during rewarming in order to discover the inridence ofarrhyth mias and whether they contribute to the mortality. - TEH 0412554 DUP050453606 S.J POCOCX, A. G. Shaper, D. Ashby, T. Delves* i ABSTRACT ! ---------------j The British Regional Heart Study is a large prospective investigation into the risk factors for ischaemic heart disease based on a cohort of men aged 40 - 59 recruited from general practice registers in 24 British towns. Blood lead concentrations and measurements of blood pressure were recorded for 7371 men at initial screening in 1978 - 80. This paper uses these data to explore the hypothesis that an elevated body lead burden may be affecting blood pressure in adults. Univariate analysis showed no association between blood lead and blood pressure, r= +.03 and +.01 for systolic and diastolic respectively. Multiple regressions of blood pressure on log (blood lead) were carried out adjusting for the following confounders: age, body mass, alcohol, smoking, observer, social class and town of residence. These estimated that doubling of blood lead is associated with an estimated increase of 1.45 mm Hg for systolic and 1.25 mm Hg for diastolic BP. The association is very weak, the study is observational and it is difficult to adjust for all relevant confounders. Therefore, it is premature to conclude that an elevated body lead burden has a causal influence on blood pressure. INTRODUCTION Considerable public concern has been shown about the effects on health of exposure to lead in adults ..and in children. In response to reports that moderate increases in blood lead concentration are associated with hypertension we present findings from a clinical survey of British middle aged men, relating measurements of blood lead concentrations to | blood pressure. | The Regional Heart Study recruited 7735 men aged 40 - 59 who were randomly sleeted from representative general practices in 24 British ' towns. Details of the selection of towns and general practices and :.e methods of screening and data collection have been reported (ref 1). j hC TWEEN TOWN COMPARISON "*he mean systolic and diastolic blood pressure were computed for men in ! m each town (around 300 men per town). There was a marked geographic variation in blood pressure. For instance, mean systolic BP ranged I *rcm 153 mm Hg in Dunferline to 136 mm Hg in Guildford. The mean blood ; tv-partment of Clinical Epidemiology & General Practice ` I:` yal Free Hospital School of Medicine, Rowland Hill Street, o.-.don NW3 2PF, United Kingdom. and `Department of chemical Pathology 8 Human Metabolism University of Southampton, U.K. TEH 0412555 i DUP050453607 N 31002.02 leaA in each tom (around 300 men per tom) showed a marked g,-,,, variation, ranging from 0.99 pmol/i in Ayr and Barrogate to 0.55in Ipswich. It was noted that six toms with the highest mean bl lead were in soft water areas and that three of these had evidence elevated levels of lead in drinking water. Thus, it seems likely.1 geographic variation in body lead burden is substantially affected the nature of water supplies. There was no association between 24 toms' mean blood lead and mean blood pressure, either systolic diastolic. COMPARISON OF INDIVIDUALS - UNIVARIATE ANALYSIS Blood pressure and blood lead concentrations were both measured in 7 men (95% of men screened). The correlation coefficients of blood le with systolic and diastolic blood pressure were r=+0.03 and r=+Q.01 respectively. The association with systolic blood pressure, althouq close to zero correlation, was significant at the 1% level. In lar" scale surveys, however, it is the magnitude of an association, not i+ significance, that is biologically important. In this same sample of! men, it was noted that blood pressure was markedly related to age and' body mass. Also alcohol intake, observer, social class, smoking andtown of residence seemed important factors to take into account (refIT?'1 Alcohol intake was of particular interest since it has been shown that': blood lead levels show a marked dose response relationship with alcohol consumption (ref 2). ALLOWANCE FOR CONFOUNDERS Multiple regression techniques have been used to relate blood pressure' (systolic or diastolic) to the natural logarithm of blood lead and also"' to the potential confounders mentioned above. The results are shorn in the following table: REGRESSIONS OF BLOOD PRESSURE ON LOG (BLOOD LEAD) Systolic BP Unadjusted adjusted for body mass, age alcohol, smoking, social class and observer adjusted also for town (Pirkle et al - NHANES XI Diastolic BP Unadjusted adjusted for body mass, age alcohol, smoking, social class and observer adjusted also for town (Pirkle et al - NHANES II Regression coefficient for log (blood lead) + 1.684 + 0.675 + 2.089 + 8.436) + 0.302 - 0.063 1.809 3.954) P-value .009 .28 .003 .46 .87 .001 TEH 0412556 DUP050453608 It can be seen that: adjustment for other personal fact$fis" {notably age, body mass and alcohol) makes the association of blood lead and blood pressure non-significant, even though this is a very large sample of middle-aged men. However, when one also adjusts for town of residence the regression coefficient for log (blood lead) becomes statistically significant both for systolic and diastolic blood pressure. The magnitude of these associations can best be expressed as follows: a doubling of blood lead is associated with an estimated increase of 1.45mm Hg for systolic BP (with 95% confidence limits 0.48, 2.42 mm Hg) and 1.25 mm Hg for diastolic BP (95% confidence limits 0.65, 1.85mm Hg). Discussion The observed associations above are much weaker than those reported by Pirkle and colleagues (ref 3). They studied blood lead and blood pressure in 564 men aged between 40 - 59 from the American National Health and Nutrition Examination Survey (NHANES II). For systolic blood pressure they found much larger regression coefficients for log (blood lead) after adjusting for age, body mass and many nutritional and biochemical variables, as indicated in the table above. However, their sample is much smaller and hence their estimation is subject to a larger random error. Also their findings relate to a selected sub group of a much larger sample of over 5,000 adults and analyses of the full information indicate a somewhat weaker association (ref 4) Adjustment for town of residence made a notable difference to our findings,. Since NHANES II was conducted in 64 locations over a period of several years it would seem important to take account of place of r' .-idence and hence time in their analyses. Since blood lead levels . in ; the incidence of hypertension have both been falling over time it ''oold be that their observed association is not present once these time Uends are accounted for. Other demographic and lifestyle factors .Vy: alcohol intake) would also merit further exploration. >i: the Regional Heart Study we have found a weak but statistically i ;nif leant positive association between blood lead and blood pressure allowance for relevant confounding variables. Since it is dif ficult to take account of all relevant confounders and the association i: weak we would consider it premature to assume that the relation- is causal and of any significance to public health. ' exploration of these findings has been reported (ref 5) and a : description will be undertaken shortly. A; 'KHOWEEDGEMENT ?:. British Regional Heart Study is supported by a programme grant from Medical Research Council. * :ftRENCES a ' Shaper et al * G Shaper et al ! J* L. Pirkle et al 4 w* R. Harlan et al $ s* <J. Pocock et al Brit Med J 283, 179 (1981) Brit Med J 284, 299 (1982) Amer J Epidemiology 121, 246 (1985) J. Amer Med Assoc 253,530 (1985) Brit Med J 289, 872 (1984 DUP050453609  "=r" BRTT&H MEDICAL JOURNAL VOLUME 291 26 OCTOBER 1985 quinine, severe hypoglycaemia was not detected (p 1169) (* (Hall and Bhattacharya, unpublished). \ Quinidine has long been recognised to be at least as good a j blood schizonticide as quinine. Recent studies in Thailand | have shown both that this drug (the D-enandomer ofquinine) is severalfold more active than quinine against P falciparum in vitro and that it may replace it for the treatment of uncomplicated infections (P Suntharasamai, S Vanijanond, T Harinasuta, et al, paper presented at 11th international congress of tropical medicine and malaria, Calgary, Canada, j 1984)." The theoretical objection to quinidine is that it | causes a greater prolongation of the Q-T interval than quinine," but it may be used to treat severe falciparum j malaria."15 Again in Thailand, White and his team recom- j mended a loading dose of quinidine of 15 mg base/kg followed by 7*5 mg base/kg every eight hours." They have not, however, done a controlled study ofintravenous quinine and intravenous quinidine. Quinine in the non-toxic doses used should normally be given for severe falciparum in- ' fections from areas other than Thailand (p 1169)2 (Hall and Bhattacharya, unpublished). Quinidine in equivalent doses should be used if quinine is not available in the hospital pharmacy. Preliminary reports suggest that an even more rapid schizonticidal response may follow the infusion of artesunate, a soluble derivative of the Chinese compound artemisinin." There were 960 patients with falciparum malaria from Africa treated in Britain between 1983 and 1984 but only one patient from Thailand (Malaria Reference Laboratory reports). Supportive management of the life threatening complica tions of severe falciparum malaria include the reduction of hyperpyrexia, control of convulsions if present, exchange blood transfusion mentioned above, haemodialysis for anuric renal failure, and endotracheal intubation with intermittent positive pressure respiration for acute pulmonary oedema or severe coma (A P Hall, unpublished). Corticosteroids are now agreed to be contraindicated in the management of falciparum malaria.' The careful maintenance of fluid balance is of vital importance, since fluid overload may precipitate acute pulmonary oedema with its serious prognosis.* Malaria caused by some strains of Pfalciparum acquired in areas where chloroquine resistance is prevalent, and es pecially in South East Asia,3 may recrudesce after initial response to quinine or, less often, to quinidine.'* This problem may be avoided by concluding treatment with either a single dose of the pyrimethamine-sulfadoxine combination Fansidar in countries where this remains effective" (Hall and Bhattacharya, unpublished) or a five to seven day course of tetracycline if resistance to this combination is present.16 Joint Director, Public Health Laboratory Service Malaria Reference Centre, London School ofHygiene and Tropical Medicine, London WCIE 7HT . Consultant Physician, Hospital for Tropica! Diseases, London NWi OPE W Pe t e r s APHa l l 1 Gyr K, Speck B, Ritz R, Cornu P, Buckner CD. ZercbraJe malaria tropical mil schwvxwu- saiieber.cfa*icMd1Pecfourfo 1974;104:1628-50. 2 Hall AP. The treactscntof severe falciparum malaria. TrantR Sac TrapMtdtfyt 1977;71:367-79. J Cbongsuphaiaisiddhi T, Sabcharoen A. Auanath P. In vivo and in viuo sesauvity of falciparum malaria to quinine in Thaictuldrcn. Arm TnpPae&atr 1981;1:21-6. 4 White NJ, Looarecsuwan S, WarrcB DA, ttal. Quinine landing doM in cerebral malaria. An] Med Hyg 1983;32:1-5. 5 WaneD DA, Looarcesuwan S, Wtrrril MJ, ei. Descnecbuooc proves deleterious w cerebral malaria. A double blind trial in 100 comatose EnflJ Med 1982;366:313-9. UK 11 wet/ u 1147 6 White NJ, Wandl DA, Ouatsmuch P, f a/. Severe fcypogiycama and faypcrimuUaa&ai ia falciparum malaria. ATE*gtJ Med 1983;391:61-6. 7 White NJ, Warrell DA, 1 mum msmi S, Qtanthavxmdi P, PtuDqrs RE Poogpaev P. PithophysiQfa>ytcalaadprnpMRM.^:i-.;frw<*Krw.i4Mfwn*l.ftiiid factatem cerebral malaria, lancer 1985^776-8. 8 White NJ, Warrell DA, Looaieesuwin S, ttaL HfpogJycnanit in fakfaanna malaria. QJMed 1982;51:508-9. 9 White NJ, Wane# DA. CEnfca! management of cbJoroquinc-resfatast PEumodium faldparum malaria in Southeast Asia. TrapDorr 1983;13:153-8. 10 White NJ, Looarecsuwan S, Warrell DA, Chongtupbaiaisiddlu T, Bunco* D, Harinasuta T. Quinidine in falciparum malaria. Lancet 198tm:1069-7l. 11 White NJ, Looarcesuwaa S, WamU DA. Quinine and quinidine: a comparison of HKG effects during die treatment ofmalaria,j CatSavau Pharmacol 19833:173-5. 12 RiHB,StahelE, Phtet JPrPriednaaumM.DatAntiarrhythaiikuinChiflid2naisAlTeRiativeindcr Bchandlusgder schweren Falciparam-Malaria. Schwa*Med Wochensckt 1983;113:254-8. 13 PhiilipaRE,WarrcUDA,WhiteNJ,LooaieesuwanS,Karbw<ngJ.lntrsveoousquiaidtneforthe * treatment of severe falciparum malaria. Clinical and pharmacokinetic studies. N EngiJ Med 1985312:1273-8. 14 LiG, Arnold KGuoX, JutnH.FuL. Randomised comparative study ofmeSoquine, qtaghaosu, and pyrunethamine-sulfadoxine in parieots with falciparum malaria, jjxncet 1984;ii:1360-l. 1$ Hail AP, Doberstyn EB, Mectaprakong V, Sookom P. Faldparum malaria cured by quinine followed by sidradorine-pyrimethamine. BrMcdJ 1975^1:15-7. 16 World Health Or^anisatioo, Aduxmcn m mahtria chemotherapy. Reportofataenaficgmtp. Genera: World Health Organisation, 1984. (WHO TeckRtpStr, No 711.) Blood lead and blood pressure Exposure to lead and its possible effect on health has long been a topic of public and scientific concern. Much of this concern has focused on the contribution of leaded petrol to the total body content of lead and the possibility that moderate rises in body lead may affect the mental develop ment of children.12 Concern has also been expressed about the link between exposure to lead and hypertension, even though no consistent relation has been shown in those who have been exposed to high concentrations of lead in an industrial setting.' The hypothesis that hypertension may be causally related to raised blood lead concentrations has been supported by the fact that the solubility of lead is increased in soft, acidic water, that in areas where the water supply is soft the mortality rate from cardiovascular disease is increased, and that hypertension is a major contributor to cardio vascular mortality. Glasgow was the centre for much of the early work on this subject, for the drinking water is extremely soft, lead plumbing systems are common, and there is a high prevalence of hypertension and a very high cardiovascular mortality rate. In 1972, in the suburb of Renfrew, 3001 men and women residents aged 45-64 were screened for hyper tension (diastolic blood pressure greater than 100 mmHg).4 Hypertension was present in 468 and the first 135 who attended a follow up clinic were matched for age and sex with 135 normotensive people randomly selected from the same population. Mean blood lead concentrations were higher in the hypertensive group, in whom concentrations greater than 2 pmol/l (41 pg/100 ml) occurred more frequently. In a paired analysis of hypertensive and normotensive men (n=74), using three categories of blood lead concentration, there were more pairs in which the hypertensive man had a higher blood lead concentration. A similar analysis for women gave results that were not statistically significant. The authors concluded that high blood pressure in the west of Scotland "is associated with high blood lead levels, which might explain the high prevalence of cardiovascular disease in the area." When this study was carried out the important influences of alcohol and smoking on blood lead were not known,' nor was there an awareness of the effects of alcohol on blood pressure.* Thus it is possible that the higher blood lead concentrations in the hypertensive subjects were due to TEH 0412558 DUP050453610 N 31002.03 R48 higher intakes ofalcohol and to the smoking that commonly accompanies drinking. Furthermore, the control group might have differed in some other way--for example, in social class or area ofRenfrew--and that may have influenced blood lead concentrations. But despite these possible drawbacks the Renfrew study is still quoted widely in support of the positive relation between blood lead concen tration and hypertension. A larger study of blood lead and blood pressure has been undertaken as part of the British Regional Heart Study. This is a prospective study ofcardiovasculardisease in Britain that aims to explain the regional variations in cardiovascular mortality and to identify the environmental and personal risk factors responsible for ischaemic heart disease.7 The study has shown striking differences in mean blood pressure between groups of men in 24 towns in England, Wales, and Scotland. These were due partly to differences in body mass index and alcohol intake, but these two factors explain only a small part of the between town variation in blood pressure. Soft water supplies are related to cardiovascular mortality rates, and towns with soft water have the highest concentra tions of lead both in the water supply and in the blood of the inhabitants.8 So, could the body lead burden (reflected in blood lead concentrations) be contributing to cardiovascular mortality, either by raising blood pressure or by some other effect on ischaemic heart disease? Blood lead concentration and blood pressure were measured in 95% of 7735 middle aged men drawn at random from general practices in the 24 towns. No positive associa tion between blood lead concentration and blood pressure was found, although among those men with blood con centrations greater than 1*8 pmol/1 (37'3 pg/100 ml) there were more hypertensive subjects than in those with lower concentrations.' Nevertheless, since this group accounted for less than 1% of the men in the study it seems unlikely that exposure to lead has an appreciable effect on blood pressure in the male population of Great Britain. These findings contrast with those reported by French workers, who studied 431 men aged 24-55 years and found a positive association between blood lead concentration and blood pressure." The association was statistically significant for men aged under 45 but not for those aged 45-55. The authors suggest that "the increase in blood lead concentration parallels the increase in blood pressure until some limit value so that such a trend is apparent only when other factors (such as age) do not competitively increase blood pressure by greater amounts." More substantial evidence of a positive association between blood lead and blood pressure comes from the United States. The second United States National Health and Nutrition Examination Survey examined 20 322 people aged 6 months to 74 years recruited from 64 different areas across the United States. For both black and white men_and women aged 12-74, blood lead concentrations were positively associated with systolic and diastolic blood pressures.11 In the younger age group (21-55 years) blood lead concentrations were significantly higher in hypertensive than in normotensive subjects but this was not the case in those aged 56-74. Some 34 other variables were also identified as having a statistically significant univariate correlation with blood pressure. Blood lead had a statistically significant relation with systolic and diastolic blood pressure in men, the magnitude of this association being about one sixth of that observed between body mass index and blood pressure. Blood lead was not independently related to blood pressure In women. Surprisingly, no alcohol effect on blood pressure was observed for men in this study, although in women BRITISH MEDICAL JOURNAL VOLUME 291 26 OCTOBER 1985 alcohol intake was independently associated with diastolic \blood pressure. The authors state that "no causal inferences should be drawn from this cross sectional survey about the blood V pressure effects of lead." Nevertheless, in a second publica tion, focusing on the results of a subgroup of564 white men aged 40-59 years, the conclusions were less guarded." The reason for selecting out this subgroup was to enable data from the United States Pooling Project and the Framingham study to be used in estimating the effects of lead on stroke, ischaemic heart disease, and all causes of mortality. Using a complex multiple regression model which included adjust ments for age, body mass index, and a large number of biochemical and nutritional measurements, a significant independent relation was shown between blood lead and both systolic and diastolic blood pressure. It needs to be emphasised, however, that values of haemoglobin, serum albumin, dietary vitamin C, potassium, riboflavin, and oleic acid were also all independendy related to either systolic or diastolic blood pressure or both, and at a statistical sig nificance equal to or greater than that achieved by blood lead. Clearly this gives cause for thought, for presumably this )_ paper could have been written to focus on any one of these variables. \/ The authors hypothesised that if lead is causally related to blood pressure, changes in blood lead concentrations will produce substantial changes in blood pressure and affect the incidence of cardiovascular events. Using the data from the pooling project and the Framingham study they calculated the "effects" of lowering blood lead concentrations on fatal and non-fatal myocardial infarction and stroke and on deaths from all causes. As the second United States National Health and Nutrition Examination Survey showed a decrease of 37% in mean blood lead concentrations during 1976-80 (from 0'81 to 0-57 pmol/1; 16*8 to 11*8 pg/100 ml) they estimated that hypertension would decrease by 17'5%, myocardial infarc tion by 4-7%, stroke by 6-7%, and deaths from all causes bv 5-5%. The implication of their calculations is that a reduction in environmental lead--derived mainly from car exhaust fumes --would save large numbers of lives as well as reducing the number of non-fatal strokes and myocardial infarctions^ Such extrapolations, however, seem simplistic.Thg'deter- minants of hypertension are complex, and the factors concerned in the development of atherosclerosis and ischaemic heart disease equally so. Thus to leap from an inconsistent relation between blood lead and blood pressure to an immediate effect of blood lead on cardiovascular mortality seems precipitate. Furthermore, a subsequent reanalysis of the same second United States National Health and Nutrition Examination Survey data on white men aged 46-65 by another group of investigators has come to a totally different conclusion (P G Gartside, letter submitted for publication). Their approach was to confine the multiple regression to a more limited set of variables which had already been identified as probable risk factors for hypertension, plus a few others that had pre viously been found to be useful. The final multiple regression analysis of 621 observations for diastolic blood pressure yielded significant associations for body mass index, coffee consumption, date of examination, and height but not for blood lead concentration (p=0`27). Similarly, systolic blood pressure was significantly related to body mass index, age, and coffee consumption but not to blood lead (p=0-18). This curious influence of coffee consumption on the relation between lead and blood pressure needs further clarification, TEH 0412559 DUP050453611 yOH MEDICAL JOURNAL. VOLUME 291 26 OCTOBER 1985 1149 / '/Since coffee is not widely regarded to be a risk factor for hypertension. (There is the current interest in the effect of i coffee on blood lipids,13 as well as hypotheses relating lipid ` metabolism to blood pressure,14 but it seems premature to attempt a reconciliation of these various themes.) In conclusion, the main evidence regarding lead and blood pressure comes from two large cross sectional surveys: the second United States National Health and Nutrition Exam ination Survey and the British Regional Heart Study. Their conflicting findings make it impossible to make a definitive statement about the effect of environmental lead on hyper tension. Further analysis of those studies may produce some rationalisation of these puzzling differences in research findings. But whatever the outcome, the issue of blood lead and hypertension is back and may remain prominent for some time. A G Sh a p e r Professor of clinical epidemiology SJPo c o c k Reader in medical statistics Department ofClinical Epidemiology and General Practice, RoyaJ Free Hospital School of Medicine, London NW32PF 3 Medical Research Council. Tkt ncuropxvchitfcgtcal effects of Uod in children: a renew of recent meareh /979-/9M. London: MRC, 1984. 2 Fococfc SJ, Ashby D. Environmental lead and children's intelligence. A review of recent epidemiological studies. The Statistician 1985;34:79-92. 3 Hunter D. Thedtteasti ofoccupation. 6th ed. London: Hodder and Stoughton, 1978:276. 4 Bcevcrs DG, Erskine E, Robertson M, et al. Blood lead and hypertension. Lancet I976;ii:l-J. 5 Shaper AG, PocockSj. Walker M,*ra/. Effects ofalcohol and smokingoa blood lead in cuddle- aged British men. BrSiedJ 1982;284:299-302. 6 Bwvcrs DC. Alcohol and hypertension. Lancet I977;u; 114-5. 7 Super AG, Pocnck SJ,'Walker M, Cohen NM.WaleCJ, Thomson AG- British regional heart study: cardiovascular risk factors in middle-aged men in 24 towns. BtMeiJ 1981;283:179-86. 8 Focock SJ, Shaper AG. Walker M, tt ai. Effects of tap water lead, water hardness, alcohol, and cigarettes on blood lead concentrations. J Epidemutl Commumiy Health 1983;37:1*7. 9 Pocock SJ, Shaper AG, Ashby D, Delves T, Whitehead TP. Blood lead concentration, blood pressure, and renal function. BrMedJ 1984;289:872-4. 10 OnsaudG. Claude JR, Moreau T, Leilouch J, Juguer B, Festy B. Blood lead concentration and blood pressure. BrMedJ 1985:290:244. 11 Harlan WR, Landis R, Schmoudcr RL, Goldstein NG, Harlan LC. Blood lead asd blood pressure.JAMA 1985^53:530-4. 32 Pirkle JL Schwartz J, Landis JR, Harland WR. The relationship between blood lead levels and blood pressure and irs cardiovascular risk implicatrons. AmJ Epidemic/1985;121:246-56. 83 Fsrde OH, Knussen SF, Amesen E, Thclle DS. The Tromsn heart study. Coffee consumption and serum liptd concentrations in men with hyperchoksterolaenila: a randomised intervention study. BrMedJ 1983;290:893-5. 14 Smith-Birharo PA. Dietary fat and blood pressure. Ann Intern Mtd 1983,98:828-31. Crohn's disease in the elderly \ i "Yourold men shall dream dreams and your young men shall i see visions" said the prophet Joel, recognising that age ; changes us and modifies our response to important events.1 / In a medical context age may affect the incidence and the type, of disease and its course, morbidity, and mortality. As the elderly population increases and the incidence of Crohn's disease rises the pattern of the disease in old people warrants examination. The few publications on the topic make a recent study from the gastrointestinal unit at Birmingham General Hospital of particular interest.' Previous studies have tended to be selective and restricted either to patients treated surgically or to those with a particular distribution of disease.*4 Crohn's disease in the elderly is rare. The Birmingham group had identified only 47 patients over the past 40 years who were aged over 60 at the time of diagnosis; they accounted for 8% of the total of over 600 patients, a proportion similar to that found by Goligher in his 500 patients.1 Epidemiological surveys based on defined popula tions rather than clinical series show that the peak incidence occurs in the third decade, after which there is a sharp fall which then tails off more gradually after 40 years of age to smaller figures in the elderly.M Some series have reported a second peak in the eighth decade, but with small numbers the finding may be spurious.* Whether the extent and distribution of disease are dif ferent in the elderly has been disputed. In che Birmingham series distal ileal disease with or without spread to the right colon was the most common form, accounting for halfofthe cases; this is true for all age groups. Distal colonic disease, however, was more common than in younger patients and accounted for 40% compared with only 6% in younger patients. Extensive colonic or small bowel disease was uncommon. This relatively high proportion of colonic disease has been recognised in epidemiological surveys,'1 and in Kyle's series from north east Scotland was most prominent in elderly women.4 The course of the disease was found to depend largely on the site of the lesion. Most of those with distal ileal disease needed laparotomy for obstruction, peritonitis, or to exclude carcinoma, but thereafter the prognosis was good. Fourteen of 22 patients with this disease distribution remained well. By contrast, patients with colonic disease rarely required surgery and were managed medically; there were only five resections in this group, and 14 of these 22 patients were also fit and well. This compares with the recently reported experience from St Mark's Hospital, where patients with colonic Crohn's disease had an accumulated probability of having bowel surgery ofnearly 40% at 10 years." One further difference was that recurrence after the initial resection--a feature of the disease in general and particularly in young patients--occurred rarely in the Birmingham series and none of the patients required a second operation. The overall standardised mortality ratio in Crohn's disease is usually quoted as twice the expected value for the general population.*,WJ The risk of death is, however, much greater for patients who develop the disease early in life, and in one series those diagnosed in the second decade had a mortality ratio 11 times the expected value." Elderly patients, how ever, have a mortality rate the same or less than the average for all patients with Crohn's disease. This may be due partly to the prominence of colonic disease, which carries a better prognosis. Diagnosis ofCrohn's disease in the elderly may be difficult because of the high prevalence of diverticular disease14 and carcinoma of the colon or caecum. Diverticular disease was present in more than halfof the patients with colonic Crohn's disease, and lack of a confident radiological diagnosis of the caecal abnormality led to laparotomy in more than a quarter of the patients with ileocaeca! disease. Recognition of the generally favourable prognosis for Crohn's disease in elderly patients, particularly those with distal colonic disease, should be of practical value to gastro enterologists, geriatricians, and surgeons. Plainly this variant is a less aggressive form of the disease than that seen in younger patients. University Hospital of Wales* Cardiff CF4 4XW Jo h n Rh o d e s Consultant physician Ja me s Ro s e Senior registrar 1 JodiUS. 2 Fafericius PJ, Gyde SN, Sboulef P, Keighley MRB, Akxander-WUlisms J, Allan RN. Crohn's disease inthe ekkrty. GW 1985;26:461-5. 3 Rusdi V, Staooowita DA. Craba'i disease in the older patient. Svg Gynecol Obsut 1980;150: 184-6. TEH 0412560 DUP050453612