hydrates normally found in the diet, such as sucrose, lactose and starch.
I . JBSKIWS. D.J.A.. L~lirs.A.R..GA..scI.L.M.A,. CocHEi. B. andAi.HhHri.K. Ci. M. XI. (1977): Decrease in postprandial insulin and glucose concentrations by guar and pcctin. A i i n . intern. Mrd. 86, 20.23. 2. JESKISS. D. J . A,. GOM. D.U.. Liims. A. R.. Ai.nI3T1, K. G. M. 34.. W0LEb'E.R. T M.S.. kWXl.1.. M. A. and HWKADAY. T. D. R. (19761: Unabsurbdhle carbohydratn and diabetes: Decreased postprandial hyperglycaemia. Lanrct, 2. I 5 1 174. 3. CAIIIIJ.. G . F.. ETZWII.I.EK, n. D. and I-KEINMI.. N.(1976): "Control" and diahetes. brew OigI J. Mud. 294. 1004. 4. JOB. I).. EXCHWEtiE. E.. i.E, P. J. (1963): lminunodssay ol'insulin with insulinantibody precipitate. Biocheni. J. 88. 137. 10. SSEUECOR. ti.1'. (1961): Statistical mcthcds. Iowa Scale College Press. Iowa.
1 1 . At.I.rx.F.M. (1920):Gross anatomic relations ofrhe pancreas and diaktes.
J. v.v/i, Afurl. 31. 381. WatlI.Qvm. M. I... ~ii.usttunsr.E. G.. MIXTON. C. R. and R~CIIARDSOX. € .N.( 1978):'~heeff~tofchain-lcngthon glucoseabsorption and therelated metabolic rcsponsc. A m r r . J. dCi. .Vim 31, 199% 2001. 13. CRAPO. P. A.. RFAVIS. Ci and Or.~ne\. J. (1977):Post prandial plasmaglucose and -insulin responses to different complex cdrhohydrdtcs. DQberrs 26, I178 11x3. 14. TRiWhI.1.. 14.. S ~ I t . I I i i i A i L .D.A,. wOl.l.VkR. D. A,. 1,kkIX. A. R.. 0.05.
t P < 0.05 otherwise P < 0.001.
Age
Alcohol
Triglycerides
0.0486 (NS) 0.1944 0.1776 0.1514 0.1049 0.3729 0.1928 N =402
164
SPICER ET A L .
analysis o f the relationship between fasting triglycerides and alcohol intake showed that the positive association was present in smokers ( r = 0,1473, P < 0.009), but not in nonsmokers ( r = 0,0447). Discussion
Blood P~-cmure In the context of New Zealand studies, agespecific mean systolic pressures in the present study are strikingly higher than those reported in the Napier7 and Milton' investigations. but d o not differ significantly at any age level from the Carterton data.'" As Simpson et u1. point out. the Milton results tend to be lower than those found in the overseas studies which they helpfully summarise. In contrast the Auckland and Carterton levels are at the upper end ofthe range of reported levels from around the world, cspecially in the 40's age group. Similarly, Auckland mean diastolic levels are notably higher, typically by about 10 mmHg, than those in Milton and Napier, but do not exceed the Carterton levels, except in the third decade. Since overseas studies have tended to identify diastolic blood pressure with the fifth phase, comparison is difficult. However, the pattern of results suggests that the Auckland levels are somewhat high by world standards. In considering the extent to which the present results are an accurate estimate of blood pressure levels in the population sampled, several points are noteworthy. Firstly, standard deviations, particularly of the mean diastolic pressures, are strikingly small. I t would be reassuring to believe that this was primarily due to the use of only one observer. However, the low variance may also reflect consistent rounding errors, for example, a tendency to prefer 95 to 90 in rounding diastolic pressures. Secondly, it should be borne in mind that the main aim of the study was to examine associations between psychological factors, notably those relating to stressful experiences, and traditional C H D risk factors. Thus, by the time subjects' blood pressure was recorded they had spent half an hour describing and discussing their stress experiences. Conceivably this procedure could have resulted in elevated blood pressures in some cases. Nevertheless, the sequence of
VOI,.
9.
NO.
2
psychological testing followed by blood pressure recording was undertaken deliberately. and without further data from a probability sample ofthe Auckland population this hypothesis must remain as speculation. Since the present results do not differ significantly from thosc in the welldesigned Carterton study, which did not include detailed psychological assessment preceding blood pressure recording. therc is no powerful reason to believe that they are grossly in error. Thirdly, t o what extent could the relatively high blood pressure levels in thc present study be due to sampling bias? Although the sample was not representative with regard to socio-economic level, no evidence could be found linking this variable with blood pressure. Age is a well known correlate of blood pressure, but the age structure of the study group accurately reflected that found in Auckland males of European origin in the 30-55 age range. Thus it seems that neither the socio-economic level nor the age characteristics of the sample provide an explanation for their blood pressure levels. A third possible biasing factor was body mass. Comparison with the Milton and Carterton data showed that the Auckland men were both taller and lighter than their similarly aged counterparts. Mean Quetelet Index scores were thus signific.antly lower in the Auckland sample than in the Milton group ( t = 5.9369, P < 0.001). Conceivably, therefore, the Auckland sample was biased with respect to body mass. However, such a bias would lead one to expect relatively low, not high, mean blood pressures in the Auckland sample. In summary. it seems that there is no obvious reason to doubt that the mean blood pressures found in this study were a reasonable estimate of those in the population sampled. Strong associations were found between age. body mass and both blood pressures. These are relationships which have been reported in other studies" and are well accepted. The apparent association between alcohol and systolic blood pressure was shown to be due to the former's correlation with body mass. This confounding effect of one variable on another illustrates well the importance of going beyond simple correlation analyses. There was some tendency for higher diastolic pressures to be associated with the presence of
APRIL
1979
COROUARY RISK FAC'TORS IN A U C KLAUD
ECG abnormalities. Given the unselected nature of the sample, this relationship was probably due to the presence o f a group of subjects with both a history of coronary heart disease and hypertension. The absence of any correlation between smoking and blood pressure replicates the Carterton finding, but contrasts with some overseas reports of an inverse association."
Choksterd Three earlier studies in Carterton2'. Dunedin4 and Milton' have provided information on the distribution of cholesterol levels in New Zealand samples. In general, the Auckland mean levels follow most closely those found in Dunedin, exhibiting no significant difference in any age group. The Carterton levels are the highest yet recorded in New Zealand, whilst the Milton figures are the lowest, except for subjects aged below 30 or above 50 where they closely resemble the Dunedin figures. It is interesting that the clearest difference to emerge is between Milton and Carterton since these are the only two studies to have used the same laboratory technique. Nye et u I . have ~ raised the issue of possible urban-rural differences in cholesterol levels. As the data now stand the urban-rural relationship is far from clear. since although the results from the two urban centres are markedly similar, they are generally sandwiched from above and below by the two rural studies. Without further data on regional differences, for example on dietary composition, including alcohol consumption, no immediate explanation for this configuration presents itself. Comparisons are made all the more difficult by the problem of sampling bias in the present study. For example, since 2201" of the sample did not attend a medical laboratory for blood testing, the characteristics of attenders and nonattenders were compared. It was found that attenders were older (r = 4.25, P < O.OOl), and had higher body mass scores ( t = 2.10, P < 0.04). On this basis alone it may be argued that the mean cholesterol level reported here is probably an overestimate of that in the target population. To place New Zealand cholesterol levels in an international context, the Busselton study in Australiaz2 indicated mean levels comparable to
165
the Carterton figures. However, a recent cooperative report from six USA centresz3 clearly shows that mean cholesterol levels in mean aged 40-59 are consistently lower than those in New Zealand samples. Bearing in mind the dangers of comparing results obtained using different laboratory techniques, it nevertheless seems likely that, as in the case of blood pressure, New Zealand cholesterol levels tend towards the upper reaches of the international range. As in other New Zealand and overseas studies, the Auckland cholesterol levels increase with age. Although they do not display the characteristic reversal beginning in the late 40's or early 5O's, there is a clear flattening of the positive trend during the fourth decade. Apart from its relationship with age, cholesterol level was also significantly associated with triglycerides, and blood pressure. This pattern of inter-relations amongst traditional CHD risk factors also indicates the importance of using multivariate statistical techniques to unravel causal relations in the etiology of the disease. Cholesterol levels also appeared to be weakly related to both body mass and cigarette smoking. Neither the Carterton nor the Dunedin study reported an association between cholesterol and smoking, but the Quetelet association was found in Carterton. In the present study the significance level of these correlations means that they must be viewed with caution.
Triglycerides Only one other study has provided information on the distribution of triglycerides in a probability sample of the New Zealand population. As in these Milton findings5, the Auckland distribution of fasting triglycerides is positively skewed. Mean levels for all age groups are consistently higher, although not markedly so, in Auckland than in Milton, peaking at 1 . 7 mmolil in the 45 -50 age group. When compared to data from the recent USA cooperative study2j, which also found positively skewed distributions, Auckland mean figures are of a similar order. In these studies mean levels ranged from 1.02-2.02 mmolil in the 40--59 age group. The fact that the Auckland levels are higher than in Milton may be partially due to the inclusion of ten cases with a definite coronary
166
SPICBR ET AL.
heart disease history, and several more with a suspected history. In this connection, Palmer et u I . ~ have ~ found that Australian coronary patients exhibit significantly higher triglycerides than did subjects from the Stockholm prospective study. However, differences in laboratory technique and the relatively small numbers in the Auckland fasting group provide a weak base on which to build any argument concerning regional differences in New Zealand. With regard to possible sampling bias only one factor could be identified which might have influenced both fasting and non-fasting triglyceride levels. I t was found that there were significantly more smokers in the lower than in the upper socio-economic levels ( X 2 = 20.03, P < 0.001). Since the lower social groups were under-represented in the sample, and since smoking and triglycerides were positively associated, it is possible that the mean triglyceride levels recorded are underestimates of those in the target population. However in the case of nonfasting triglycerides another factor may have biased the results in the opposite direction. Subjects who provided blood samples were significantly heavier than those who did not, a fact which could have given an inflated estimate of the population mean triglyceride level. On the basis of the available data, and given the general similarity between the Auckland findings and those from Milton and the USA, there are no grounds for suspecting serious sampling bias in the triglycerides results. Examination of the fasting and non-fasting distributions in the present study shows the expected higher mean, variance and range in the latter. It is however noteworthy that much of the difference in mean scores can be attributed to a small number of subjects in the non-fasting group with levels in excess of 5 mmol/l. Unlike the Milton investigation the present results show no relationship between fasting triglycerides and age. Nor is age associated with non-fasting triglyceride levels in Auckland. This lack of correlation could well be due to both the relatively small numbers in the fasting group and to the limited age range. The Milton data indicate that during the 30-55 age range the association between triglycerides and age changes direction. Also the US cooperative
__
VOL.
9,
NO.
2
project reports a tendency for fasting triglyceride levels to decline from age 40 onwards. This phenomenon, occurring in a relatively small data set, could conceivably result in the disappearance of a significant correlation. Cigarette smoking was related to both fasting and non-fasting triglycerides, while the apparent effect of alcohol intake on fasting levels was clearly shown to be spurious. The smoking relationship is of some interest since it suggests that requiring patients to fast may not be a sufficient precaution in attempting to measure their “true” triglyceride levels. The finding that body mass was related to non-fasting but not to fasting triglyceride levels may also be of relevance to this point. It may suggest that the body mass of those subjects who exhibit marked lability of triglycerides in response to food intake is significantly different to that of those subjects who do not show such lability. Uratc> The range of urate levels recorded in Auckland was strikingly similar to that found in Carterton (0.19-0.58 m m o l ~ l ) . Carterton ~~ age-specific mean levels of 0.3-0.39 mmol/l were also very comparable to the central tendency of the Auckland data. Conceivably the Auckland mean levels were biased upwards since subjects who provided a blood sample were heavier than those who did not, and body mass is a significant correlate of serum urate level. But without data from another community sample this must remain as speculation. A recent study of serum urate levels in Japanese men in Hawaii2’ found a very similar unimodal distribution with a mean of 0-36mmolil. These authors note that their results are intermediate in mean levels between those found in Caucasians and in Filipinos or South Pacific Islanders. Thus, once again New Zealand levels appear to lie at the upper end of the range found in US and European studies. In common with the Carterton data the Auckland results exhibit no linear relationship between urate and age. As in many other studies in New Zealand and overseas, body mass was found to be an important correlate of urate level. Evidence concerning the relationship between habitual alcohol intake and serum urate is mixed, but on balance affirmative.26 Our findings sup-
APKlL
1979
COKOhARY RISK FAClOKS IN AUCKLAND
port those from Carterton and Hawaii in displaying a strong positive correlation between serum urate and alcohol intake which is independent of body mass. Our findings on the relationship between serum urate and blood pressure, notably a weak association with diastolic pressure. add to the conflicting evidence reviewed by Yano P t ~ 1In . general the present results on serum urate and lipids are consistent with those from population studies which show that. whilst triglycerides are an important correlate of serum urate, cholesterol and serum urate are not significantly related.
Glucose There is little evidence in the literature on the distribution of glucose levels assessed without a glucose tolerance test in the New Zealand population. Abernethy et ~11.’’ recently reported a random blood sugar diabetes detection survey in Christchurch, where they found a mean nonfasting level o f 4 . 9 mmol:’/(88 mg/dl) in a group of over 3000 volunteers. However, given the heterogeneity of this group no meaningful comparisons with the present results are possible. A striking feature of the Auckland results is the similarity in the distributions of fasting and non-fasting levels. Although the range and variance o f the non-fating distribution are significantly greater, the mean is not significantly higher than that in the fasting group. I t is traditional in epidemiokgical studies to prefer measures of fasting glucose levels on the assumption that glucose lability is a common phenomenon. However, our results suggest that the majority of subjects maintain glucose levels within a relatively narrow range. Thus, for the purposes of an epidemiological study, assessing subjects in a non-fasting state may well provide a sufficiently accurate estimate of the distribution of blood glucose levels. One reservation t o this argument concerns a possible source of bias in the fasting glucose results. As in the case of triglycerides. fasting glucose was related to body mass, and subjects who provided blood samples were notably heavier than those who did not. Thus, the mean fasting glucose level may have been biased upwards, bringing it closer to the non-fasting mean level and thereby removing the
167
expected difference. N o other possible source of sampling error could be identified in the glucose findings. Only two variables, body mass and ECG arrhythmias, were correlated with glucose level, and these only in the fasting group. Once again it is possible that the correlation between arrhyth~ ~ mias and glucose level is due to the presence of subjects with both a coronary history and asymptomatic hyperglycaemia, a suspected CH D risk factor.2 An alternative hypothesis is that both levels may indicate marked sympathetic arousal in a particular sub-group in response to the psychological stresses of ECG recording and giving a blood sample. However, once again the weakness of the correlations with glucose level necessitates caution in their acceptance and interpretation.
Cigarette and Alcohol Consumption The percentage of subjects who were regular smokers (47”/) is not too dissimilar from the 1976 New Zealand census data which showed that 42-5”/,, of males in the 30-59 age group smoke regularly.” However, it is probable that the Auckland figure is an underestimate for the population sampled. As noted in the section on triglycerides, smokers were more common in the lower social classes, a group which were underrepresented in the sample. Our estimates of the percentages of men who had never smoked and who had successfully given up were also strikingly close to the census data for men in this age group. With respect to the actual number of cigarettes consumed each day, well over three quarters of those who smoke cigarettes consume in excess of ten a day, a pattern of consumption even more marked in the census data. Since there is a clear dose-response relationship between risk of CHD and the number of cigarettes smoked, this pattern of consumption suggests that there is great scope for preventative action in New Zealand at least in terms of encouraging smokers to reduce their intake. Although habitual alcohol intake is not established as an independent risk factor in CHD, it is instructive to examine the pattern of subjects’ consumption, if only because of its apparent relationship to body mass, smoking and serum
168
SPICER ET AL.
VOL.
9,
NO.
2
urate levels. Few epidemiological data have been published in New Zealand. Direct comparisons with other studies such as Carterton2’ are precluded because of the use of different measuring procedures. However the distribution of alcohol consumption both in Auckland and Carterton is consistent with the general proposition that high intake is restricted to a very small proportion of the population, the great majority consuming less than 300g of alcohol in an average week. This figure should perhaps be viewed with some caution as a possible underestimate of that in the population sampled. Although no relationship was found between alcohol intake and socio-economic level, the association of both these variables with cigarette smoking, and the under-representation of the lower social classes in the sample do raise the possibility of bias.
men. Once again this hypothesis must await further study. The results presented in this paper suggest that in the group studied the prevalence pattern of traditional risk factors was comparable to that found in other New Zealand and overseas investigations. In general, there appears to be a tendency for New Zealand values to lie at the upper end of the ranges reported in the world literature, a fact to be considered in conjunction with New Zealand’s relatively high incidence of CHD.
ECG Findings With three exceptions the ECG findings were similar to those reported in the Seven Countries Studies.” T wave abnormalities were comparatively infrequent in Auckland, a finding which may be partly explained by the younger age groups studied. AV conduction abnormalities, as defined by Minnesota Code 6.3, were notably more frequent in Auckland. There was a striking tendency for the records in this category to exhibit only a slightly prolonged PR interval (0.22-0.24 sec), and a rather slow heart rate (50-60 beats/min). The latter point may well provide an explanation for the apparently high prevalence of AV conduction abnormalities. Of related interest is the fact that in only four of the 17 samples reported in the Seven Countries Study did the prevalence of sinus bradycardia approximate or exceed the Auckland figure. There is no ready explanation for the high number of subjects with relatively slow heart rates, but a contributory factor may be the apparent popularity of physical fitness programmes amongst Auckland men. Unfortunately no attempt was made to assess physical activity in this study, so this hypothesis remains untested. It is also possible that there is a positive association between socio-economic level and physical fitness, in which case the sample would have contained an unusually high number of fit
References
Acknowledgements The authors would like to thank: Ms Gaynor Cave11 and Ms Ngaire Barrack for their assistance in the data collection; Ms Rosemary Rees for her analysis of the ECG records; members of the Biostatistics Unit in the Department of Community Health for their guidance in the statistical analyses, and Ms Lynda Burgess for secretarial assistance.
I
National Heart Foundation of New Zealand (1976): Cbronarv Heorr Diacwe A Progrrs.5 Reporr
2. S ~ A M L I.. ~ RBFRKSON. , D. M and LINDBERG,H. A. (1972). Risk factors: Their role in the etiology and pathogenesis of the atherosclerotic diseases. In: The Pufhogmne.srha j Alherosclerosrs. Wissler, R. W. and Geer, J. C. (edsj. Williams and Wilkins. Baltimore. D. 41 3. PR1on.l.A. M. (1974):CardiovascularepidemiologyinNewZealandandthe Pacific, M.Z. med. J . 80, 245. 4 HUNILR,J D and WONG,L C. K. (1961): Plasmacholesterol levels in New Zealand. Observations in I000 urhan males. Brrr med. J. 2, 486. 5. NYE, E. R., SU!HFRLANU, W. H. F., LARKING, P. W. and SPLARS. G . F. S. (1977): Blood lipids and lipoproteins in a rural New Zealand population, Ausf. .V.Z. J. Wed 1, 134. 6. SIMPSON.F. 0..WAAI.-MAYYING.H. I., B O L L I ,andSman$O. ~ F S (1978): The Milton Survey: Blood pressure and heart rate, N L.med. J. (In press). 7. ( ~ I S T M A S .B. W. (1977):A survey of blood pressure levels of an adult Ne* Zealand urhan oooulatmn. Nanier 1973. N.Z. rned. J . 86. 369. 8. GRIEVE. R J (1977): Distribution of various coronary nsk factors in an urhan general practice, N . Z . r n ~ dJ. 86, 178. 9. SCorr,P. J. ilY67): Serum low-density lipoprotein levels in a New Zealand population sample, Clin chim. Arlo 15, 449. 10. SPICLR,I.. M~LEOIJ.W. R.. O’BRIES, K P and SWTT, P. J (1978): Pwchoswial correlates of coronary heart disease: Distnbutions and associations with traditional nsk Eactora in a random sample of Auckland men. (Suhmittcd for publication.) I I ROSE, Ci. A and BLACKBURN. H. (IY68J Cardiovascular survey methods. World Health Organisation, Geneva, p. 137 12 LFVINE. J., MOROENSTERN, S and VLASTILICA, D. (1967): Automation in analytical chemistry. Technicon Symposia. 13. FOSTER, L. B and DUNN, R. T. (1973): Stable reagents for determination of serum triglycerides by a colorimetric Hantzsch condensation method, Chn Chem. 199, 338 14. LfVFR, .M., Pow~1.1.. J C., KILLIP,M. and SMALL,C W. (1973): A comparison of 4-hydroxybenzoic and hydrazidc (PAHBAH) with other rcagents for the detcrmination of glucose, J Lab L l m . Med. 82:4, 649. 15. NFISON, J. W and BATRA, K K. (1975): Simplified automated method for determination of urinary or serum uric acid, based on reduction of ferricphenanthroline complex. Chn. Chem. 21.1, 125 I6 NIE. N . H , HULL,C H., JENKINS, I. G..STEINHRENYtR, K. and BENT,D. H. (1975): Statistical package for the social sciences. Second Editron, McGrawHill, Pie, York 17. Census of population and dwellings. provisional national statistics (1976) Department of Statistics, Wellington, New Zealand 18. EI.I.TY, W. B. and IRVING,J. C (1972): A swio-economic Index for New Zealand based on levels ofeducation and income from the 1966 census, N.Z J. Educ Stud 1, 153 19. EVAYS. J. A ,PRIOR,1. A M. AND M~RRISOV. R . B. 1. (1969):The Carterton Study 3 Blood pressures of a sample of New Zealand European adults, N .Z . mcd. J 69, 146 I SBL4cKHUnY, , H. W., VAh k L H E Y , F. s. P.. B U Z I ~ A , 20 KFYS,A,. A R A ~ ~ NC., R , DJORDJtVK’. B S , DONTAS, A S., FtUANZA, F , K A R V O W t 3 , M. I., KIMIJRA.N.,LEKOS,D.. MONTI,M., PLDDU.V and TAYLOR, H. L (1966) Epidemiological studies related to coronary heart disease: Characteristics of men aged 40-59 in seven countries, A r f a med .mind.. Suppl. 460.
..
APRIL
1979
, PRIOK. I A M and CGOKr. N. 4 . Serum cholesterol leve!s of a %ampleof N Z nwd .I. 69. 346 22. CLRZOW.D. H., C ~ L L F ' ~K. J . MCCAI1.. M. G Srwiiot-sk'.N S and W ~Biimw. I T A. (1969) Hcalthanddiscasc~naruralcommunify. 4Western Australian Sludy. ,4101. J . Sn.31, 281. 23 CASTI:I,LI.W. P , C(X1t'FR. ( i K , DOYLE. 1 T . ~ARCIA-~'.&l.YltRI. M.. GOR~OK. T , HAMS. C.. HLLIt Y . S. R , KGAK A . KWIIMAK. U M ( , G L F ~ D. and Virtc, W J . (1977) Distribution of tnglycende and rota1 LVL and HDL cholesterol In several populations: A cooperatiw lipoproicin phcnoiypmg study, J . ihron. D b 30, 147. 24. PA1MI.R. A J . BLACKFT. R. 8. and LErLARlHAFPlh, B. (19??1. Hyperhpidacmld in a group of coronary subject, in Sydney. , W P ~J Aurf Special Suppl. 2. 19
21
169
CORONARY RISK FACTORS IX AUCKLANII G., I'KI~R.I..4 M.arid I.(lln~rs~i.;.R B I (19691 *rum iinc acid lwei) o f a sarnplc dNeu Lcdldnd adul
bVAh.7. J
.
26 YAW),K , RHOAI)?.G G nod K . ~ . ~ AAY ,(1977) Epidcniiology 01 serum uric acid among 8000 Japanrsr r2rncncnn men in Hawaii. J &on. 111s 30.
M H . Ahimb. C . Bt,%wh.,D. W.. T%iLOK. A rmdom blood sugar diahetea detection \UTY
Y.
123.
Aust. N.Z.
J. Med. (19791. 9. DD. 160- 173
Exercise Testing in Congenital Aortic Stenosis K. F. Hossack" and G. H. Neilsont
From the Cardiac Investigation Unit, The Prince Charles Hospital, Brisbane, Queensland
SUmmary:
Exercise testing in congenital aortic stenosis. K. F. Hossack and G. H.Neilson, Aust. N.Z. J. Med. 1979, 9, pp. 169-173.
Treadmill exercise testing was performed on 23 patients with isolated congenital aortic stenosis. It was found that eight patients, with a negative exercise test and a normal rise in systolic blood pressure on exercise, had gradients less than 45 mmHg. Fifteen patients with gradients greater than 50 mmHg had positive exercise tests and their systolic blood pressure did not rise normally. It is suggested that left heart studies to measure aortic valve gradients can safely be postponed if an exercise test is negative. Serial exercise testing of such patients will give an indication when left heart catheterization should be considered.
The clinical assessment of the severity of congenital aortic stenosis, in the absence of significant electrocardiographic changes, can be very d i f f i c ~ 1 t .In l ~middle aged patients the quality of the carotid pulse is a very reliable index of the *Cardiology Registrar. *Medical Director. Correspondence: Dr. G. H. Neilson, Cardiac Investigation Unit, The Prince Charles Hospital, Rode Road, Chermside, Queensland 4032 Australia Accepted for publication: 13 November, 1978
severity of the lesion', ",but in young children the carotid pulse can be normal despite significant aortic stenosis. Using phonocardiography and analysis of the carotid upstroke', severe aortic stenosis was invariably present if the three derived parameters of ejection time index, maximal rate of rise of carotid pulse, and time of the peak of the systolic murmur all fell outside the set limits. However, significant aortic stenosis can exist in the presence of only one parameter being abnormal. In a collaborative study' a formula for predicting aortic valve gradient was proposed. The formula involved the intensity of the murmur and the size of the R and Q waves in lead V, on the electrocardiogram. Eighty per cent of the patients with a measured aortic valve gradient of less than 50 mmHg were accurately predicted, but with measured gradients greater than 80 mmHg, 20% were predicted as having gradients of less than 50 mmHg. There have been several reports3y9 which indicate that a positive exercise test correlates well with an aortic valve gradient of greater than 50 mmHg. We decided to investigate the value of treadmill exercise testing in a group of young patients with congenital aortic stenosis who had normal ST segments and upright T waves in the resting electrocardiograms.
I . JBSKIWS. D.J.A.. L~lirs.A.R..GA..scI.L.M.A,. CocHEi. B. andAi.HhHri.K. Ci. M. XI. (1977): Decrease in postprandial insulin and glucose concentrations by guar and pcctin. A i i n . intern. Mrd. 86, 20.23. 2. JESKISS. D. J . A,. GOM. D.U.. Liims. A. R.. Ai.nI3T1, K. G. M. 34.. W0LEb'E.R. T M.S.. kWXl.1.. M. A. and HWKADAY. T. D. R. (19761: Unabsurbdhle carbohydratn and diabetes: Decreased postprandial hyperglycaemia. Lanrct, 2. I 5 1 174. 3. CAIIIIJ.. G . F.. ETZWII.I.EK, n. D. and I-KEINMI.. N.(1976): "Control" and diahetes. brew OigI J. Mud. 294. 1004. 4. JOB. I).. EXCHWEtiE. E.. i.E, P. J. (1963): lminunodssay ol'insulin with insulinantibody precipitate. Biocheni. J. 88. 137. 10. SSEUECOR. ti.1'. (1961): Statistical mcthcds. Iowa Scale College Press. Iowa.
1 1 . At.I.rx.F.M. (1920):Gross anatomic relations ofrhe pancreas and diaktes.
J. v.v/i, Afurl. 31. 381. WatlI.Qvm. M. I... ~ii.usttunsr.E. G.. MIXTON. C. R. and R~CIIARDSOX. € .N.( 1978):'~heeff~tofchain-lcngthon glucoseabsorption and therelated metabolic rcsponsc. A m r r . J. dCi. .Vim 31, 199% 2001. 13. CRAPO. P. A.. RFAVIS. Ci and Or.~ne\. J. (1977):Post prandial plasmaglucose and -insulin responses to different complex cdrhohydrdtcs. DQberrs 26, I178 11x3. 14. TRiWhI.1.. 14.. S ~ I t . I I i i i A i L .D.A,. wOl.l.VkR. D. A,. 1,kkIX. A. R.. 0.05.
t P < 0.05 otherwise P < 0.001.
Age
Alcohol
Triglycerides
0.0486 (NS) 0.1944 0.1776 0.1514 0.1049 0.3729 0.1928 N =402
164
SPICER ET A L .
analysis o f the relationship between fasting triglycerides and alcohol intake showed that the positive association was present in smokers ( r = 0,1473, P < 0.009), but not in nonsmokers ( r = 0,0447). Discussion
Blood P~-cmure In the context of New Zealand studies, agespecific mean systolic pressures in the present study are strikingly higher than those reported in the Napier7 and Milton' investigations. but d o not differ significantly at any age level from the Carterton data.'" As Simpson et u1. point out. the Milton results tend to be lower than those found in the overseas studies which they helpfully summarise. In contrast the Auckland and Carterton levels are at the upper end ofthe range of reported levels from around the world, cspecially in the 40's age group. Similarly, Auckland mean diastolic levels are notably higher, typically by about 10 mmHg, than those in Milton and Napier, but do not exceed the Carterton levels, except in the third decade. Since overseas studies have tended to identify diastolic blood pressure with the fifth phase, comparison is difficult. However, the pattern of results suggests that the Auckland levels are somewhat high by world standards. In considering the extent to which the present results are an accurate estimate of blood pressure levels in the population sampled, several points are noteworthy. Firstly, standard deviations, particularly of the mean diastolic pressures, are strikingly small. I t would be reassuring to believe that this was primarily due to the use of only one observer. However, the low variance may also reflect consistent rounding errors, for example, a tendency to prefer 95 to 90 in rounding diastolic pressures. Secondly, it should be borne in mind that the main aim of the study was to examine associations between psychological factors, notably those relating to stressful experiences, and traditional C H D risk factors. Thus, by the time subjects' blood pressure was recorded they had spent half an hour describing and discussing their stress experiences. Conceivably this procedure could have resulted in elevated blood pressures in some cases. Nevertheless, the sequence of
VOI,.
9.
NO.
2
psychological testing followed by blood pressure recording was undertaken deliberately. and without further data from a probability sample ofthe Auckland population this hypothesis must remain as speculation. Since the present results do not differ significantly from thosc in the welldesigned Carterton study, which did not include detailed psychological assessment preceding blood pressure recording. therc is no powerful reason to believe that they are grossly in error. Thirdly, t o what extent could the relatively high blood pressure levels in thc present study be due to sampling bias? Although the sample was not representative with regard to socio-economic level, no evidence could be found linking this variable with blood pressure. Age is a well known correlate of blood pressure, but the age structure of the study group accurately reflected that found in Auckland males of European origin in the 30-55 age range. Thus it seems that neither the socio-economic level nor the age characteristics of the sample provide an explanation for their blood pressure levels. A third possible biasing factor was body mass. Comparison with the Milton and Carterton data showed that the Auckland men were both taller and lighter than their similarly aged counterparts. Mean Quetelet Index scores were thus signific.antly lower in the Auckland sample than in the Milton group ( t = 5.9369, P < 0.001). Conceivably, therefore, the Auckland sample was biased with respect to body mass. However, such a bias would lead one to expect relatively low, not high, mean blood pressures in the Auckland sample. In summary. it seems that there is no obvious reason to doubt that the mean blood pressures found in this study were a reasonable estimate of those in the population sampled. Strong associations were found between age. body mass and both blood pressures. These are relationships which have been reported in other studies" and are well accepted. The apparent association between alcohol and systolic blood pressure was shown to be due to the former's correlation with body mass. This confounding effect of one variable on another illustrates well the importance of going beyond simple correlation analyses. There was some tendency for higher diastolic pressures to be associated with the presence of
APRIL
1979
COROUARY RISK FAC'TORS IN A U C KLAUD
ECG abnormalities. Given the unselected nature of the sample, this relationship was probably due to the presence o f a group of subjects with both a history of coronary heart disease and hypertension. The absence of any correlation between smoking and blood pressure replicates the Carterton finding, but contrasts with some overseas reports of an inverse association."
Choksterd Three earlier studies in Carterton2'. Dunedin4 and Milton' have provided information on the distribution of cholesterol levels in New Zealand samples. In general, the Auckland mean levels follow most closely those found in Dunedin, exhibiting no significant difference in any age group. The Carterton levels are the highest yet recorded in New Zealand, whilst the Milton figures are the lowest, except for subjects aged below 30 or above 50 where they closely resemble the Dunedin figures. It is interesting that the clearest difference to emerge is between Milton and Carterton since these are the only two studies to have used the same laboratory technique. Nye et u I . have ~ raised the issue of possible urban-rural differences in cholesterol levels. As the data now stand the urban-rural relationship is far from clear. since although the results from the two urban centres are markedly similar, they are generally sandwiched from above and below by the two rural studies. Without further data on regional differences, for example on dietary composition, including alcohol consumption, no immediate explanation for this configuration presents itself. Comparisons are made all the more difficult by the problem of sampling bias in the present study. For example, since 2201" of the sample did not attend a medical laboratory for blood testing, the characteristics of attenders and nonattenders were compared. It was found that attenders were older (r = 4.25, P < O.OOl), and had higher body mass scores ( t = 2.10, P < 0.04). On this basis alone it may be argued that the mean cholesterol level reported here is probably an overestimate of that in the target population. To place New Zealand cholesterol levels in an international context, the Busselton study in Australiaz2 indicated mean levels comparable to
165
the Carterton figures. However, a recent cooperative report from six USA centresz3 clearly shows that mean cholesterol levels in mean aged 40-59 are consistently lower than those in New Zealand samples. Bearing in mind the dangers of comparing results obtained using different laboratory techniques, it nevertheless seems likely that, as in the case of blood pressure, New Zealand cholesterol levels tend towards the upper reaches of the international range. As in other New Zealand and overseas studies, the Auckland cholesterol levels increase with age. Although they do not display the characteristic reversal beginning in the late 40's or early 5O's, there is a clear flattening of the positive trend during the fourth decade. Apart from its relationship with age, cholesterol level was also significantly associated with triglycerides, and blood pressure. This pattern of inter-relations amongst traditional CHD risk factors also indicates the importance of using multivariate statistical techniques to unravel causal relations in the etiology of the disease. Cholesterol levels also appeared to be weakly related to both body mass and cigarette smoking. Neither the Carterton nor the Dunedin study reported an association between cholesterol and smoking, but the Quetelet association was found in Carterton. In the present study the significance level of these correlations means that they must be viewed with caution.
Triglycerides Only one other study has provided information on the distribution of triglycerides in a probability sample of the New Zealand population. As in these Milton findings5, the Auckland distribution of fasting triglycerides is positively skewed. Mean levels for all age groups are consistently higher, although not markedly so, in Auckland than in Milton, peaking at 1 . 7 mmolil in the 45 -50 age group. When compared to data from the recent USA cooperative study2j, which also found positively skewed distributions, Auckland mean figures are of a similar order. In these studies mean levels ranged from 1.02-2.02 mmolil in the 40--59 age group. The fact that the Auckland levels are higher than in Milton may be partially due to the inclusion of ten cases with a definite coronary
166
SPICBR ET AL.
heart disease history, and several more with a suspected history. In this connection, Palmer et u I . ~ have ~ found that Australian coronary patients exhibit significantly higher triglycerides than did subjects from the Stockholm prospective study. However, differences in laboratory technique and the relatively small numbers in the Auckland fasting group provide a weak base on which to build any argument concerning regional differences in New Zealand. With regard to possible sampling bias only one factor could be identified which might have influenced both fasting and non-fasting triglyceride levels. I t was found that there were significantly more smokers in the lower than in the upper socio-economic levels ( X 2 = 20.03, P < 0.001). Since the lower social groups were under-represented in the sample, and since smoking and triglycerides were positively associated, it is possible that the mean triglyceride levels recorded are underestimates of those in the target population. However in the case of nonfasting triglycerides another factor may have biased the results in the opposite direction. Subjects who provided blood samples were significantly heavier than those who did not, a fact which could have given an inflated estimate of the population mean triglyceride level. On the basis of the available data, and given the general similarity between the Auckland findings and those from Milton and the USA, there are no grounds for suspecting serious sampling bias in the triglycerides results. Examination of the fasting and non-fasting distributions in the present study shows the expected higher mean, variance and range in the latter. It is however noteworthy that much of the difference in mean scores can be attributed to a small number of subjects in the non-fasting group with levels in excess of 5 mmol/l. Unlike the Milton investigation the present results show no relationship between fasting triglycerides and age. Nor is age associated with non-fasting triglyceride levels in Auckland. This lack of correlation could well be due to both the relatively small numbers in the fasting group and to the limited age range. The Milton data indicate that during the 30-55 age range the association between triglycerides and age changes direction. Also the US cooperative
__
VOL.
9,
NO.
2
project reports a tendency for fasting triglyceride levels to decline from age 40 onwards. This phenomenon, occurring in a relatively small data set, could conceivably result in the disappearance of a significant correlation. Cigarette smoking was related to both fasting and non-fasting triglycerides, while the apparent effect of alcohol intake on fasting levels was clearly shown to be spurious. The smoking relationship is of some interest since it suggests that requiring patients to fast may not be a sufficient precaution in attempting to measure their “true” triglyceride levels. The finding that body mass was related to non-fasting but not to fasting triglyceride levels may also be of relevance to this point. It may suggest that the body mass of those subjects who exhibit marked lability of triglycerides in response to food intake is significantly different to that of those subjects who do not show such lability. Uratc> The range of urate levels recorded in Auckland was strikingly similar to that found in Carterton (0.19-0.58 m m o l ~ l ) . Carterton ~~ age-specific mean levels of 0.3-0.39 mmol/l were also very comparable to the central tendency of the Auckland data. Conceivably the Auckland mean levels were biased upwards since subjects who provided a blood sample were heavier than those who did not, and body mass is a significant correlate of serum urate level. But without data from another community sample this must remain as speculation. A recent study of serum urate levels in Japanese men in Hawaii2’ found a very similar unimodal distribution with a mean of 0-36mmolil. These authors note that their results are intermediate in mean levels between those found in Caucasians and in Filipinos or South Pacific Islanders. Thus, once again New Zealand levels appear to lie at the upper end of the range found in US and European studies. In common with the Carterton data the Auckland results exhibit no linear relationship between urate and age. As in many other studies in New Zealand and overseas, body mass was found to be an important correlate of urate level. Evidence concerning the relationship between habitual alcohol intake and serum urate is mixed, but on balance affirmative.26 Our findings sup-
APKlL
1979
COKOhARY RISK FAClOKS IN AUCKLAND
port those from Carterton and Hawaii in displaying a strong positive correlation between serum urate and alcohol intake which is independent of body mass. Our findings on the relationship between serum urate and blood pressure, notably a weak association with diastolic pressure. add to the conflicting evidence reviewed by Yano P t ~ 1In . general the present results on serum urate and lipids are consistent with those from population studies which show that. whilst triglycerides are an important correlate of serum urate, cholesterol and serum urate are not significantly related.
Glucose There is little evidence in the literature on the distribution of glucose levels assessed without a glucose tolerance test in the New Zealand population. Abernethy et ~11.’’ recently reported a random blood sugar diabetes detection survey in Christchurch, where they found a mean nonfasting level o f 4 . 9 mmol:’/(88 mg/dl) in a group of over 3000 volunteers. However, given the heterogeneity of this group no meaningful comparisons with the present results are possible. A striking feature of the Auckland results is the similarity in the distributions of fasting and non-fasting levels. Although the range and variance o f the non-fating distribution are significantly greater, the mean is not significantly higher than that in the fasting group. I t is traditional in epidemiokgical studies to prefer measures of fasting glucose levels on the assumption that glucose lability is a common phenomenon. However, our results suggest that the majority of subjects maintain glucose levels within a relatively narrow range. Thus, for the purposes of an epidemiological study, assessing subjects in a non-fasting state may well provide a sufficiently accurate estimate of the distribution of blood glucose levels. One reservation t o this argument concerns a possible source of bias in the fasting glucose results. As in the case of triglycerides. fasting glucose was related to body mass, and subjects who provided blood samples were notably heavier than those who did not. Thus, the mean fasting glucose level may have been biased upwards, bringing it closer to the non-fasting mean level and thereby removing the
167
expected difference. N o other possible source of sampling error could be identified in the glucose findings. Only two variables, body mass and ECG arrhythmias, were correlated with glucose level, and these only in the fasting group. Once again it is possible that the correlation between arrhyth~ ~ mias and glucose level is due to the presence of subjects with both a coronary history and asymptomatic hyperglycaemia, a suspected CH D risk factor.2 An alternative hypothesis is that both levels may indicate marked sympathetic arousal in a particular sub-group in response to the psychological stresses of ECG recording and giving a blood sample. However, once again the weakness of the correlations with glucose level necessitates caution in their acceptance and interpretation.
Cigarette and Alcohol Consumption The percentage of subjects who were regular smokers (47”/) is not too dissimilar from the 1976 New Zealand census data which showed that 42-5”/,, of males in the 30-59 age group smoke regularly.” However, it is probable that the Auckland figure is an underestimate for the population sampled. As noted in the section on triglycerides, smokers were more common in the lower social classes, a group which were underrepresented in the sample. Our estimates of the percentages of men who had never smoked and who had successfully given up were also strikingly close to the census data for men in this age group. With respect to the actual number of cigarettes consumed each day, well over three quarters of those who smoke cigarettes consume in excess of ten a day, a pattern of consumption even more marked in the census data. Since there is a clear dose-response relationship between risk of CHD and the number of cigarettes smoked, this pattern of consumption suggests that there is great scope for preventative action in New Zealand at least in terms of encouraging smokers to reduce their intake. Although habitual alcohol intake is not established as an independent risk factor in CHD, it is instructive to examine the pattern of subjects’ consumption, if only because of its apparent relationship to body mass, smoking and serum
168
SPICER ET AL.
VOL.
9,
NO.
2
urate levels. Few epidemiological data have been published in New Zealand. Direct comparisons with other studies such as Carterton2’ are precluded because of the use of different measuring procedures. However the distribution of alcohol consumption both in Auckland and Carterton is consistent with the general proposition that high intake is restricted to a very small proportion of the population, the great majority consuming less than 300g of alcohol in an average week. This figure should perhaps be viewed with some caution as a possible underestimate of that in the population sampled. Although no relationship was found between alcohol intake and socio-economic level, the association of both these variables with cigarette smoking, and the under-representation of the lower social classes in the sample do raise the possibility of bias.
men. Once again this hypothesis must await further study. The results presented in this paper suggest that in the group studied the prevalence pattern of traditional risk factors was comparable to that found in other New Zealand and overseas investigations. In general, there appears to be a tendency for New Zealand values to lie at the upper end of the ranges reported in the world literature, a fact to be considered in conjunction with New Zealand’s relatively high incidence of CHD.
ECG Findings With three exceptions the ECG findings were similar to those reported in the Seven Countries Studies.” T wave abnormalities were comparatively infrequent in Auckland, a finding which may be partly explained by the younger age groups studied. AV conduction abnormalities, as defined by Minnesota Code 6.3, were notably more frequent in Auckland. There was a striking tendency for the records in this category to exhibit only a slightly prolonged PR interval (0.22-0.24 sec), and a rather slow heart rate (50-60 beats/min). The latter point may well provide an explanation for the apparently high prevalence of AV conduction abnormalities. Of related interest is the fact that in only four of the 17 samples reported in the Seven Countries Study did the prevalence of sinus bradycardia approximate or exceed the Auckland figure. There is no ready explanation for the high number of subjects with relatively slow heart rates, but a contributory factor may be the apparent popularity of physical fitness programmes amongst Auckland men. Unfortunately no attempt was made to assess physical activity in this study, so this hypothesis remains untested. It is also possible that there is a positive association between socio-economic level and physical fitness, in which case the sample would have contained an unusually high number of fit
References
Acknowledgements The authors would like to thank: Ms Gaynor Cave11 and Ms Ngaire Barrack for their assistance in the data collection; Ms Rosemary Rees for her analysis of the ECG records; members of the Biostatistics Unit in the Department of Community Health for their guidance in the statistical analyses, and Ms Lynda Burgess for secretarial assistance.
I
National Heart Foundation of New Zealand (1976): Cbronarv Heorr Diacwe A Progrrs.5 Reporr
2. S ~ A M L I.. ~ RBFRKSON. , D. M and LINDBERG,H. A. (1972). Risk factors: Their role in the etiology and pathogenesis of the atherosclerotic diseases. In: The Pufhogmne.srha j Alherosclerosrs. Wissler, R. W. and Geer, J. C. (edsj. Williams and Wilkins. Baltimore. D. 41 3. PR1on.l.A. M. (1974):CardiovascularepidemiologyinNewZealandandthe Pacific, M.Z. med. J . 80, 245. 4 HUNILR,J D and WONG,L C. K. (1961): Plasmacholesterol levels in New Zealand. Observations in I000 urhan males. Brrr med. J. 2, 486. 5. NYE, E. R., SU!HFRLANU, W. H. F., LARKING, P. W. and SPLARS. G . F. S. (1977): Blood lipids and lipoproteins in a rural New Zealand population, Ausf. .V.Z. J. Wed 1, 134. 6. SIMPSON.F. 0..WAAI.-MAYYING.H. I., B O L L I ,andSman$O. ~ F S (1978): The Milton Survey: Blood pressure and heart rate, N L.med. J. (In press). 7. ( ~ I S T M A S .B. W. (1977):A survey of blood pressure levels of an adult Ne* Zealand urhan oooulatmn. Nanier 1973. N.Z. rned. J . 86. 369. 8. GRIEVE. R J (1977): Distribution of various coronary nsk factors in an urhan general practice, N . Z . r n ~ dJ. 86, 178. 9. SCorr,P. J. ilY67): Serum low-density lipoprotein levels in a New Zealand population sample, Clin chim. Arlo 15, 449. 10. SPICLR,I.. M~LEOIJ.W. R.. O’BRIES, K P and SWTT, P. J (1978): Pwchoswial correlates of coronary heart disease: Distnbutions and associations with traditional nsk Eactora in a random sample of Auckland men. (Suhmittcd for publication.) I I ROSE, Ci. A and BLACKBURN. H. (IY68J Cardiovascular survey methods. World Health Organisation, Geneva, p. 137 12 LFVINE. J., MOROENSTERN, S and VLASTILICA, D. (1967): Automation in analytical chemistry. Technicon Symposia. 13. FOSTER, L. B and DUNN, R. T. (1973): Stable reagents for determination of serum triglycerides by a colorimetric Hantzsch condensation method, Chn Chem. 199, 338 14. LfVFR, .M., Pow~1.1.. J C., KILLIP,M. and SMALL,C W. (1973): A comparison of 4-hydroxybenzoic and hydrazidc (PAHBAH) with other rcagents for the detcrmination of glucose, J Lab L l m . Med. 82:4, 649. 15. NFISON, J. W and BATRA, K K. (1975): Simplified automated method for determination of urinary or serum uric acid, based on reduction of ferricphenanthroline complex. Chn. Chem. 21.1, 125 I6 NIE. N . H , HULL,C H., JENKINS, I. G..STEINHRENYtR, K. and BENT,D. H. (1975): Statistical package for the social sciences. Second Editron, McGrawHill, Pie, York 17. Census of population and dwellings. provisional national statistics (1976) Department of Statistics, Wellington, New Zealand 18. EI.I.TY, W. B. and IRVING,J. C (1972): A swio-economic Index for New Zealand based on levels ofeducation and income from the 1966 census, N.Z J. Educ Stud 1, 153 19. EVAYS. J. A ,PRIOR,1. A M. AND M~RRISOV. R . B. 1. (1969):The Carterton Study 3 Blood pressures of a sample of New Zealand European adults, N .Z . mcd. J 69, 146 I SBL4cKHUnY, , H. W., VAh k L H E Y , F. s. P.. B U Z I ~ A , 20 KFYS,A,. A R A ~ ~ NC., R , DJORDJtVK’. B S , DONTAS, A S., FtUANZA, F , K A R V O W t 3 , M. I., KIMIJRA.N.,LEKOS,D.. MONTI,M., PLDDU.V and TAYLOR, H. L (1966) Epidemiological studies related to coronary heart disease: Characteristics of men aged 40-59 in seven countries, A r f a med .mind.. Suppl. 460.
..
APRIL
1979
, PRIOK. I A M and CGOKr. N. 4 . Serum cholesterol leve!s of a %ampleof N Z nwd .I. 69. 346 22. CLRZOW.D. H., C ~ L L F ' ~K. J . MCCAI1.. M. G Srwiiot-sk'.N S and W ~Biimw. I T A. (1969) Hcalthanddiscasc~naruralcommunify. 4Western Australian Sludy. ,4101. J . Sn.31, 281. 23 CASTI:I,LI.W. P , C(X1t'FR. ( i K , DOYLE. 1 T . ~ARCIA-~'.&l.YltRI. M.. GOR~OK. T , HAMS. C.. HLLIt Y . S. R , KGAK A . KWIIMAK. U M ( , G L F ~ D. and Virtc, W J . (1977) Distribution of tnglycende and rota1 LVL and HDL cholesterol In several populations: A cooperatiw lipoproicin phcnoiypmg study, J . ihron. D b 30, 147. 24. PA1MI.R. A J . BLACKFT. R. 8. and LErLARlHAFPlh, B. (19??1. Hyperhpidacmld in a group of coronary subject, in Sydney. , W P ~J Aurf Special Suppl. 2. 19
21
169
CORONARY RISK FACTORS IX AUCKLANII G., I'KI~R.I..4 M.arid I.(lln~rs~i.;.R B I (19691 *rum iinc acid lwei) o f a sarnplc dNeu Lcdldnd adul
bVAh.7. J
.
26 YAW),K , RHOAI)?.G G nod K . ~ . ~ AAY ,(1977) Epidcniiology 01 serum uric acid among 8000 Japanrsr r2rncncnn men in Hawaii. J &on. 111s 30.
M H . Ahimb. C . Bt,%wh.,D. W.. T%iLOK. A rmdom blood sugar diahetea detection \UTY
Y.
123.
Aust. N.Z.
J. Med. (19791. 9. DD. 160- 173
Exercise Testing in Congenital Aortic Stenosis K. F. Hossack" and G. H. Neilsont
From the Cardiac Investigation Unit, The Prince Charles Hospital, Brisbane, Queensland
SUmmary:
Exercise testing in congenital aortic stenosis. K. F. Hossack and G. H.Neilson, Aust. N.Z. J. Med. 1979, 9, pp. 169-173.
Treadmill exercise testing was performed on 23 patients with isolated congenital aortic stenosis. It was found that eight patients, with a negative exercise test and a normal rise in systolic blood pressure on exercise, had gradients less than 45 mmHg. Fifteen patients with gradients greater than 50 mmHg had positive exercise tests and their systolic blood pressure did not rise normally. It is suggested that left heart studies to measure aortic valve gradients can safely be postponed if an exercise test is negative. Serial exercise testing of such patients will give an indication when left heart catheterization should be considered.
The clinical assessment of the severity of congenital aortic stenosis, in the absence of significant electrocardiographic changes, can be very d i f f i c ~ 1 t .In l ~middle aged patients the quality of the carotid pulse is a very reliable index of the *Cardiology Registrar. *Medical Director. Correspondence: Dr. G. H. Neilson, Cardiac Investigation Unit, The Prince Charles Hospital, Rode Road, Chermside, Queensland 4032 Australia Accepted for publication: 13 November, 1978
severity of the lesion', ",but in young children the carotid pulse can be normal despite significant aortic stenosis. Using phonocardiography and analysis of the carotid upstroke', severe aortic stenosis was invariably present if the three derived parameters of ejection time index, maximal rate of rise of carotid pulse, and time of the peak of the systolic murmur all fell outside the set limits. However, significant aortic stenosis can exist in the presence of only one parameter being abnormal. In a collaborative study' a formula for predicting aortic valve gradient was proposed. The formula involved the intensity of the murmur and the size of the R and Q waves in lead V, on the electrocardiogram. Eighty per cent of the patients with a measured aortic valve gradient of less than 50 mmHg were accurately predicted, but with measured gradients greater than 80 mmHg, 20% were predicted as having gradients of less than 50 mmHg. There have been several reports3y9 which indicate that a positive exercise test correlates well with an aortic valve gradient of greater than 50 mmHg. We decided to investigate the value of treadmill exercise testing in a group of young patients with congenital aortic stenosis who had normal ST segments and upright T waves in the resting electrocardiograms.
