Document ypba5vv8n5LnLXXw10aGJkVy2

t The Lancet Saturday 3 July 1976 BLOOD-LEAD AND HYPERTENSION D. G.Beev er s Eil e e n Er s k in e Mo ir a Ro b e r t s o n atcal Research Council Blood Pressure Unit, \X'tstcrn Infirmary, Glasgow A. D. Be a t t ie A. Go l d b e r g B. C. Ca mp b e l l M. R. Mo o r e ersity Department ofMateria Mediae, SiobhtU Hospital, V. M. Ha w t h o r n e varment of Community Medicine, University of Glasgow, Ruefull Hospital nmary Blood and lap-water lead levels were ex amined in 135 hypertensives and 135 and sex matched normoiensives. Among male hypersives there was a significant excess of cases with high ' 3 lead levels and a similar but statistically non-signi- trend was found amongst female hypertensives. A -live correla; -n was found between blood-lead and > err lead. 1' is concluded that in the West of Scot` ntgh blood-pressure is associated with high bloodJ levels, which might explain the high prevalence of J.'.'vascular disease in the area. Introduction Hy per t en s io n and cardiovascular diseases are mmon in areas with soft drinking-water supplies.1-4 -e acidic nature of soft water leads to increased plumvlvency,5 so that in areas like the West of Scotland here the drinking water ts extremely soft and is fre_:mly delivered and stored in lead plumbing systems gh lead levels may occur in the water.6 It is not known -ether these high levels of water lead may explain the ;h prevalence of hypertension and cardiovascular dis-w in these areas.1 * Several workers have demon 's'td a significant positive correlation between the si content of the drinking water supply and blood-lead . ris.`910 Furthermore, blood-lead levels (greater than conventional upper acceptable limit of 2 umol/1 were `.nd in 4 oul of 71 subjects (5-6'ft : studied in urban s-gow,- none of whom had clinical evidence of lead .'Onmg.6 In that study there were 6 cases with diasto. Mood-pressures oi 95 mmHg or more who had a cr.er mean blood-level or 1-61 umol/1 than in normo- :ie cases studied (mean blood-lead 1-09 umol/1 /. No statistical significance could be attached to these small numbers. To investigate further the possible association of high blood-level levels with hypertension we have studied an unselected group of hypertensives and normotensive controls, detected at a screening survey in Renfrew, Scotland.11 Methods Renfrew is an urban burgh on th?omskins of Greater Glas gow. Much of the mains water-supply is delivered in lead pipes and in many of the older tenement blocks in the area, storage tanks for drinking-water are also lead-lined. There are no local industries with particular occupational hazards from lead exposure. In 1972, an epidemiological screening survey was conducted amongst residents aged 45-64 years, and diastolic blood-pres sures of 100 mmHg or more were found in 21-25? of.the popu lation studied." After screening, a follow-up clinic was estab lished to investigate and manage the hypertensive patients and carry out clinical research into hypertension." The study of blood-lead was confined to the first 135 cases to attend the clinic whose diastolic blood-pressures were 100 mmHg or greater at screening or who had had hypertension diagnosed previously by their family doctors. In all cases this level of hypertension was confirmed at the clinic. The 135 hyperten sives (74 males, 61 females) were compared with 135 age and sex matched normotensive subjects selected randomly from the same population. The diastolic blood-pressure Df controls was persistently less than 90 mmHg. No subject was occupationally exposed to lead. Blood-pressures at screening and in the clinic were mea sured using the London School of Hygiene and Tropical Medi cine sphyg momanomeier,15 diastolic blood-pressures being taken at the point of final disappearance of the diastolic sound (Korotkoff v). A 10 ml specimen of blood was taken for lead analysis into a heparinised plastic tube and subjects were pro vided with a 250 ml lead-free container and asked to return with the first sample of kitchen cold tap-water taken in the morning. Only 90 hypertensives and 119 normotensives pro vided tap-water specimens, although blood-lead levels were available in all cases. Blood and water lead levels were mea sured by flamelcss atomic absorption spectrophotometry (Pcrkin-Elmer 306: HGA 72). Data on smoking habits, obesity index, and plasma-creatinine concentrations were available in all cases from the preliminary screening examination. Results Men and women were analysed separately. Fig. 1 and 2 show the frequency distribution curves for blood-lead concentrations. The mean blood-lead concentrations for hypertensive and normotensive men were 1-31 (0-54 7975 t The Lancet Ltd, 1976 2 THE LASCET.ntV 3, 1976 BLOOD LEAD tirnot/l Fig. 2--Distribution curves for blood-lead levels in hypertensive women- Fig. 1--Distribution curves for Wood'Iced levels is sormoteoslve nd hypertensive men. s.D.) and 1-14 (0-38 s.D.) pmol/1, and for hypertensive and normoiensive women 1-09 (0-54 s.D.) and 0-94 (0-39 s.D.) linioV). High blood-lead levels (greater than 2 pmolA) were found in 8-2% of hypertensives and 1-55? of normotensives (table i). Significance tests aimed at delecting a simple shift of distribution, such as Student's ( or nonparametric sign tests, were unlikely to detect the signifi cance of an excess of raised blood-levels in one group compared with another. Instead, formal tests of signifi cance were based on table it. Amongst male subjects there were 38 (35 + 3+0) pairs of normotensives and hypertensives with blood-lead levels within the same range. If there was nn association between blood-lead and raised blood-pressure, then the normoiensive in about half of the remaining 36 pairs would be expected .to have a higher blood-lead than the hypertensive. In fact this happened in only 11 pairs. Reference to the binomial distribution yielded a two-tailed significance level of p<0-05. Thus the observed association between blood-lead and hypertension amongst men is statistically significant. Amongst women, these tests showed a simi lar trend but did not reach statistical significance. The groups studied were selected in terms of age and sex but, as seen from table t, there appeared to be a tend ency amongst hypertensives and normotensives for high blood-lead levels to occur in older subjects, although TABLE I-- BLOOO-LEAD, AGE, BLOOD-PRESSURE, SMOKING, OBESITY, ANO PLASMA-CREATININE IN SUBJECTS STUDIES Male Xtv Mean ape -vr* Mean dia-nolic bluud-pressurr tmmHg' Mean plasma-creatinine jiinull Mean *, obesity index `t smokers *4 7 water lead 5^100 (ig/l t'o'iiile. No. Mean yf Mean diaMolu Hx\l pto-Mirv imulU W.mu plJM'u vrVJimiiu: nmol V Mean*, index *r smokers *r water lend5* lOOpg'l Normoiensive, with blood-lead:* Norma* * Borderline 57 55-5 (i 5-8) 75-7 i 6-5) 100-7 (2)6-7) -MiilM) 45-6 12*5 $3 56-2 l 5-9'- > V,- f v SS-4 .ilV -3-5 43*4 111 16 54*5 (+ 5-2} 79-4 (+ 6-9) (+13-3) *2*9 (+ 8*9) 43-6 61-5 7 59*0 ( 6*5) -J5.0 4.*' <L-*2 *.** -1-8 1+10-9' 42*8 57*1 High 1 60 70 88-4 -- 16*8 100 0 1 61 #4 1P6 - 10-2 0 IDO Normal Hypertensive, with blood-lead: Borderline 46 55-5(15*4 ) 19 34-1 ( 5-8 211-0 ,i 12-7 109 0 (i 9-3) KT-6 (127-9; + 9-9(216*1) 28-2 29-7 104-7 HH.|; * 10-3 (itt-T. 36-8 30-B 45 56-4 . 6*0 n:*: sii>* 14 S6-4'6-l- 100 a - - ; sf-4 =r** * If--" ,122-7 21-7 21-9 1 1 ] SS-4 -r*" 19-9 + 29-9 21-4 27-3 High 9 5"-" ; 7-4 111-6 * 12*6 106 ! tr20 7 -9-3 slS-6' 6( -6 4".9 t-J-S * ].4 - 5 ; ?J ICt 1 r 1'-' -f* f - 25-4. 50 100 Nt'mial--'ft-1 -4 fjmol/!. l5,m3irrlwit=rl-5--1*9 j j t u o VI. Highs '2-0umol/l Division factor for conversion frotnttood-lead pmald 10 pg/100 ml=0-D483. Figures in parentheses--s.o. Obcsttv indcxs',' variation correned for height) from mean for population in Renfrew. EH 0350327 DUP050298964 INCET, Jl'LV 3, 1976 3 TABU 11" :O.MPAmSON OT BPOOP-LEAD 1.EVELS IN' FA11MD SOBUOTlMIkl AVU HYP1RHMS1VE MLN hypertension, possibly the mechanism being effects of lead on renal function. An alternative hypothesis is that otenMse O-K'jJ wN IKpenemne blood-lead levels Normal 15 ynsol 1 Borderline 1 5-1-9 unsold High . '2-0 utlUll 1 7 oial hypertensives tend to develop high blood-lead concentra tion, either by drinking more water or by chronic reten tion of lead, possibly due to reduced renal clearance as part of a hypertensive nephrosclerosis. The lack of as sociation of raised blood-lead with raised plasma-crea 2 ^mol 1 l 0 0 l tinine does not support either of these possibilities. A me l-9ftTnol1 . ].j IP 3 3 16 third hypothesis is that chronic ingestion of lead in drinking-water results in higher frequency of raised 1; 16 6 5~ blood-lead levels with advancing age, and since hyper 46 19 9 74 tension also occurs more commonly in older subjects, there might only be a chance association between these in the narrow 20-vear age range there was no corre- two age-related indices. Suppression of renin release, as n between blood-lead and age (male hypertensives, is found in up to 30(? of hypertensives, has also been 0-05). No correlation was found between the level reported in cases of chronic plumbism, even in the ,ood-pressure and blood-lead amongst hypertensives absence of hypertension.20 21 Renin suppression is also es, r=--0-02; females, r=0 02) and no correlations age-related, there being an inverse correlation between : found with plasma-creatinine and blood-lead renin and age.22-24 This suppression has been taken as )-06). 6 hypertensives had evidence of renal failure, evidence of the development of hypertensive nephros i raised plasma-creatinine concentrations; 1 had clerosis.25 26 However, in the age-matched normoiensive >nic pyelonephritis and 1 had hypertension following group studied here, high blood-lead levels were not loch-Schoeniein purpura. None of these 6 had high common, and thus the third hypothesis is less likely. >d-lead levels. It therefore remains possible that the added combina Chilst the proportion of hypertensives who were tion of soft drinking-water and a high water-lead con se was greater than the normotensives, no relalion- centration, as seen in the West of Scotland, may be a ' was found between blood-lead and obesity (male factor to explain the high prevalence of hypertension >enensives, r^O-05). It is of interest that there were and cardiovascular disease in the area. Further studies re non-smokers amongst the hypertensives than the of blood-pressure and blood-lead in a larger population motensives. are necessary to resolve this point. tVatcr-lead concentrations greater than 100 ug/1 48 (imol/1) (the World Health Organisation's upper .'eptable limit) were found in 26-6% of hypertensives d 20-1 % of normotensives. There was a relationship tween water lead and blood-lead. Significant correla,ns were found between water and blood lead tongst normotensives .(males, r=0-372, P<0-0); We are grateful for advice from Mr D. A. McLaren, Department of Statistics, University of Glasgow, and for the support of the Renfrew shire King Edward Memorial Trust, and a grant from the European Economic Community and the Scottish Home and Health Department. Requests for reprints should be addresses to D. G. B., M.R.C. Blood Pressure Unit, Western Infirmary, Glasgow G11 6NT. males, r=0-359, p<0-01) but not amongst hyperten- REFERENCES es(males, r= 0-065;females, r=0-120). Discussion Workers in a lead-related industry have been 1. Stitt. F. U., Clayton, M. D., Crawford, D. G,, Morns, J. hi. Lancet, 1973, ,122. 2. Elwood, P. C., Abernethy,M., Morton, M. ibid 1974, u, 1470. 3. Schroedcr, H. A. 7. Am. med.Ass 1966,195, 81. 4. Blorck, G., Bosinim, H., Winront, A. Acta med. scand 1965,178, 239. Bt.tncJ.J. 1972, ii, 491. ported to have an excess death-rate from cerebrovasjlai disease,14 although Cramer and Dahlberg15 were 5. Crawford, M. D., Morris, J. N. Lancet, 1967, u, 1087. 6. Beattie, A. D., Moore, M. R., Devcnav, VT, 1.. Miller. A. R., Goldberg, A. flr. med.J. 1972, ii, 491. ruble to find a raised prevalence of hypertension. Also i chronic plumbism, other changes have been reported, -.eluding intranuclear inclusion bodies in the proximal cnal tubular cells, although renal function may be pared.16 17 However, after prolonged low-grade expojre, renal failure, hyperuricsemia, and gout may -cur.1* The present study confirms the data of Beattie et al.6 "towing a possible association of chronic low-grade \posure to high-water-lead with hypertension. Elwood r. al.J in Wales confirmed a negative association between - ater hardness and cardiovascular mortality and found i weakly positive correlation between water lead and ..'tal cardiovascular mortality. However, these workers rve since failed to find consistently higher water-lead .vets in subjects dying of cardiovascular disease com- ired with other causes.19 There are several possible hypotheses to explain the ndtng of an association between blood-lead and raised bod-pressure. Chronic low-grade exposure to lead in nnking-water may in some way lead to development of 7. Hawthorne, V. M., Gillis, C. R., Larimer. A. R., Calvert, V. R., Walker, T.J. >*/. 1969, iv, 651. ! Registrar General ^Scotland). Annual Report Part 1, H.M. Stationers Office Edinburgh, 1974. 9 AJJis, G.,Moore,M. R.Nature, 1974,252, 120 10. Coveil, B. HUh Bull., Scon. Home Hitk Dtp. 1975, 33, 114. 11. Hawthorne. V. M., Greaves, D. A., Beevers, D. G. Be. Med J. 1974, m,600. 12. Beevers, 1). G., Duncan, 5., Nelson, C. S., Padneld, P. I.. Postgrad, mcd. y. (in the press i. 13. Rose, G. A., Holland,\X. IX*., Crowley, E A. Lancet, 1964, i, 296. 14. Dmguall-Fordyce, I., Lane, R. E. Br.J. Jrtd. Mod. 1963,20, 313. 15 Cramer, K., Dahlberg, L. ibid. 1966,23,101. 16 Cramer, K., Goyer, R. A., Jagenburg. R., WiHon, M.H. ibid. 1974.31,113. 17. Weeds*n, R. C. Am.y. Med 1975, 5,630. 16. Beattie, A. D., Dagg. J. H., Goldberg, A., Wang, l., Ronald, J. flr med. J. 1972, ji,4R8. 19 J.hviHvd, P. C.. St 3-cgcr, A S., Moore, F., Morton, M Lancet. 1976, i, 748. 2U SanJucaJ, H H., Michelakis. A M., Temple, 1. E Archs envirctn. Hhh IV'O, 2D, 356 21 McMlistcr, R G., Michelakis, A M., Sandstead, H. H Archs intern. Med. 19"l, 172, V19. 22. Schalekamp, M A., Kraus*. X. H., Scbalrkamp-Kuvken. M. P A., Kohlers, G., Birkenhager, XK H. Chn.Sn 1971,41,219 23. PaJfield. P. L., Beevcn, 17. G., Brown, } j., Danes, D 1... Lever. A K. Robertson, J. 1. S.. Schalekamp, M., 1 rev, M. Lancet, IV'i, i, 548 24. l uck. M. 1... Williams. C H., Cam, J. I\, Sullivan, J M., Dluhv, R. C. Am J. Cardial. 1973,32,63*' 25. Brown, J. J., Lever, A- K, Robertson, J. 1. S., Schalekamp, M. A. Lance:. I974.ii.321 26 Swales, J. D. ibid. 1975, t, 75. DUP0502 98965