1127
schizophrenic inpatients consuming 2800 kcal daily, it would have been helpful if Sacks et al had provided full baseline data since the usual diet, including 3 eggs per week, fell rather neatly between none and 7 eggs per week.
DETECTION OF THYROID NODULARITY
SIR,-It would have been of interest if Dr Hay and his colleagues p 916), in their report of the prevalence of thyroid nodularity in hyperparathyroidism (41-6%) detected by high-
(April 21,
resolution real-time ultrasonography (HRS), had stated their intraoperative findings. In our Billings Hospital series1 of 124 patients, 16-9% had a palpably abnormal thyroid preoperatively, but the intraoperative detection of visible or palpable thyroid nodularity rose to 54%. In our necropsy controls (n = 69) 46 - 4% had visible or palpable nodularity. Since these figures are not statistically different from that for HRS it
seems
that HRS does
not
offer
We were not given separate baseline values for LDL and VLDL but together they added to 112 nig/dI, as against 111 mg/dl with no eggs in the diet and 121 mg/dl with an egg a day. Thus although there is a difference in the total LDL cholesterol of 11 mg/dl between no eggs and 7 eggs per week the difference appears only on adding an egg a day to the baseline mean of 3 a week, whilst subtracting those 3 eggs appears to produce no effect at all. The Keys formula prediction for total plasma cholesterol changes reflects this:
a
significant advantage over the experienced thyroid surgeon or pathologist, respectively, in the detection of visible or palpable thyroid abnormality. Moreover, in our necropsy controls 53-6% had visibly and palpably normal thyroids, yet the prevalence of microscopic disease was 48-7%, compared with 5007o in Mortensen’s larger series.2The overall prevalence of macroscopic and microscopic thyroid abnormality in our controls was 72-5%. This would indicate that the clinical, scintigraphic, and HRS detection of thyroid nodules is still imperfect.
Eggs/wk Baselzne -0
King’s College Hospital, London SE5 9RS
E. G. LEVER
Billings Hospital, University of Chicago
E. L. KAPLAN
MATERNAL SERUM AFP AS SCREENING TEST FOR DOWN SYNDROME
SIR,-Dr Cuckle and his colleagues (April 28, p 926) calculated that, by using the criteria they suggest for screening for Down syndrome, 40% of pregnancies with Down syndrome and 6 - 807o of unaffected pregnancies would be selected. On the figures they cite for England and Wales for 1980 they would have detected only 321 Down syndrome pregnancies and selected 44 404 normal pregnancies in that year. This alarming total number of false positives with consequent parental distress, taken in conjunction with the number of miscarriages that would result from the amniocenteses performed (444 at the generally accepted miscarriage rate of 1 %, although this may be lower in specialist ceritres), surely leads one to the answer to the question they pose indirectly in their discussion. It is not acceptable to offer amniocentesis to over 6% of pregnant women on the basis of this proposed screening policy.
Eggs Authority, Eridge Road, Tunbridge Wells,
W.T.HOULSBY
Sheffield S10 2TH
EGGS AND CHOLESTEROL
SiR,-It seems to be the open season for eggs. I presume that the simple souls who cannot comprehend a more complex
same
explanation for a raised serum cholesterol than too much in the food find it easiest to identify cholesterol with eggs. Even so the data presented by Dr Sacks and others (March 24, p 647) are unworthy of your columns. Their title suggests that eating eggs raises LDLcholesterol. Their data do not support that assertion nor indeed any relation between eggs in the diet and cholesterol in the blood. In their discussion they claim that the Keys equation "should apply to changes in LDL-cholesterol". This formula stemmed from three experiments in schizophrenic men whose total blood cholesterol was measured. Since Sacks et al have total cholesterol values we should perhaps consider them: Z=(418 mg cholesterol/day)-’ (1 - 9161000 kcal/day) Z=(97 mg cholesterol/day)-(1761 x 1000 kcal/day) soZ;=I4-7, Z, =7-4, andA Cho)=I-5 (14-7-7-4) =10-95. Actual figure from study 6. 8:i: 16.8. =
The changes in total cholesterol should have been the same as LDL but were not. Since young lactovegetarian students consuming 1800 kcal daily compare uneasily with middle-aged
2
Keys pointed out that "very large cholesterol intakes may not have more effect than the highest intakes considered" by him, which were 1500 mg per day. Thus if Sacks and his colleagues were selecting the most likely group in which to "prove" the relation between dietary intake and plasma cholesterol, they did the best they could in choosing young lactovegetarians on low cholesterol intakes. It is for them a pity that the data were not up to it. From my reading of these data, the Keys formula, under rigorous double-blind crossover conditions, underestimated the change produced by adding 4 eggs a week to the baseline 3 by 55% and overestimated the effect of removing the 3 baseline eggs by 100%. Set in context, the changes measured appear to be no more than random fluctuations, as one might expect from the evidence of many other studies attempting to relate dietary cholesterol to plasma cholesterol. I see no reason to alter my view that, in healthy subjects of normal body weight and taking adequate, regular physical exercise the number of eggs eaten as part of a varied diet has no significant effect on the serum cholesterol or, therefore, on the ischaemic heart disease risk. Union House,
Department of Paediatrics, University of Sheffield, Children’s Hospital,
1. Lever
Basehne- 7
Dietary cholesterol (mg/day) Keys prediction (mg/dl) Measured change (mg/dl)
EG, Refetoff S, Strauss FH, Nguyen M, Kaplan EL. Coexisting thyroid and parathyroid disease-are they related? Surgery 1983; 94: 893-900. Mortensen JD, Woolner LB, Bennett WA. Gross and microscopic findings in clinically normal thyroid glands. J Clin Endocrinol 1955; 15: 1270-80
KentTN48HF
ALEXANDER L. MACNAIR
SiR,-Dr Sacks and colleagues claim that their study has confirmed a similar one by Roberts et al,l whose abstract ends: "The feeding of whole egg in a double-blind study in outpatients eating their customary diets had a hypercholesterolemic effect compared to a cholesterol-free diet." The "customary diet" of Roberts’ sixteen volunteers was remarkably low in cholesterol (236 mg/day ±112 SD). Eight volunteers, aged 22-61, were fed an additional 492 (±119) mg cholesterol per day in the form of a whole egg product for 4 weeks; they were then switched for a second 4 week period to a cholesterolfree substitute that looked and tasted the same. This exchange resulted in a daily cholesterol intake of 196 ±112 mg, an amount similar to the intake on the customary diet. The other eight were fed the same two diets in reverse order. Blood lipids were measured three times on each regimen but only once during the baseline control period. Each volunteer served as his/her own control. On average, total plasma cholesterol levels were 9% higher on addition of the whole egg product and 2% lower on the cholesterol-free egg substitute (p<0-01) than they were at baseline. However, individually only seven of the sixteen volunteers exhibited a pattern of change consistent with the paper’s conclusion (see figure, open circles). The other nine responded differently-three with no changes at all from baseline levels on either the higher or the lower 1
Roberts SL, McMurry MP, Connor WE Does egg feeding (ie dietary cholesterol) affect plasma cholesterol levels in humans? The results ofa double-blind study. Am J Clin Nutr 1981, 34: 2092-99.
1128
testing the effects of changes in cholesterol intake in hypercholesteraemic males were reported in abstract in 1982:4 in forty-one trials lasting 12 weeks, we found significant increases in plasma total cholesterol in only 15%, with evidence of physiological compensation for increased cholesterol intake in 71%. The same heterogeneity has been observed by us in thirty additional exchanges (unpublished). Rockefeller
University,
New York, NY 10021, USA
change in plasma total cholesterol concentrations in 16 volunteers, comparing effects of high (+)) and low ( - )cholesterol intakes. =77 who responded consistently with increases on the higher intake and decreases on the lower intake; tt=4 4 who responded anomalously; =3 3 who did not respond to either regimen; and. 2 whose levels fell below baseline on the low-cholesterol intake. All subjects consumed low-cholesterol diets I during the baseline control period. Source: Roberts et al.
%
=
cholesterol intake, four with unexpected changes (no rise in cholesterol on the higher intake but a decrease on the lower or an increase on the higher intake but no decline when the cholesterolfree regimen was substituted), and two with significant declines when fed the cholesterol-free product and an increase when the whole egg product was fed. Perhaps the habitual cholesterol intakes of these two volunteers were higher than the average for the entire group; if so, it would be reasonable to include them with the first seven. But however we assign these two volunteers about half of those participating did not respond as claimed in the abstract of Roberts’ paper. The design of Sacks and colleagues’ experiment was similar to that of Roberts. Blood samples were drawn only three times-on the habitual diet with three eggs per week and on the no-egg and one-egg regimens. On average, there was a 12% increase in LDL-cholesterol and a 9% increase in apolipoprotein B when the cholesterol intake was shifted from the no-egg diet to the one-egg regimen. Sacks et al did not find (as Roberts et al did) a change in plasma total cholesterol, which remained at 175 mg/dl when the no-egg regimen was substituted for the habitual baseline diet, despite a decrease in intake from 251 to 97 mg/day. Although individual data were not cited, Sacks et al noted considerable variability in their seventeen volunteers’ responses to the two test regimens: changes in LDL-cholesterol varied from 22% to + 42% on the higher cholesterol intake (thirteen rises, one no change, three decreases), while apo-B changes varied from - 14% to + 38% (eleven rises, two no change, four decreases). Thus out of seventeen volunteers four to six did not conform to the expected -
pattern.
Any heterogeneity of responsiveness to diet and/or drug interventions will be glossed over if data are presented as averages. People are not pure-bred laboratory animals.2 Our studies over the past 15 years have aimed at measuring this heterogeneity and have demonstrated how inappropriate it is to make broad recommendations for general population about reducing cholesterol intake.3If we could differentiate between responders and non-responders we might be able to identify those who may benefit from a lowered cholesterol intake and those for whom such advice is irrelevant, even counterproductive. Our own findings on 2. McNamara 3. Ahrens EH 4. McNamara
DJ Diet and hyperlipidemia. Arch Intern Med 1982; 42: 1121-24. Jr. Dietary fats and coronary heart disease. Lancet 1979, ii: 1345-48. DJ, Kolb R, Parker TS, et al. Heterogeneity of control of cholesterol homeostasis in man Arteriosclerosis 1982; 2: 422a
E. H. AHRENS,
JR
SIR,-The appearance of still another report showing a small hypercholesteraemic effect of feeding an egg a day for 17 days to vegetarians is sure to confuse your readers. What are they to tell their patients about diet and heart disease? The feeding trials showed that one egg a day raised serum level of cholesterol by 3’99 mg/dl. Is that a meaningful change? And what about the other extravagantly expensive studies which showed that neither drugs, niacin, clofibrate,2 cholestyramine,3diet,4or even a combination of several of those agents with treatment of smoking5and hypertensionhave had a beneficial effect on the subsequent health of the subjects. The argument is that even small reductions ofblood cholesterol must be important because they seem to agree with Keys’ prediction.Sir Robert Platt’s remark8-"squeezing the last drop of blood out of a foregone conclusion" -characteris es the diet/heart advocates. Sentencing patients to a lifetime of dietary abstinence with such tenuous evidence of efficacy is neither ethically nor scientifically sound. The antics of the dietists reminds me of a story written in 1881 by Nikolai Leskov.9 He wrote about the fabulous metal workers of Tula, a community 40 miles south of Moscow. The metal work of these craftsmen was so fine and their reputation so widespread that even British metal workers became jealous. With great skill the British craftsmen constructed a metal flea, life size and complete in all details. They sent the marvel to Tula as a challenge. After a time the flea was returned, fitted with horseshoes. I think the Boston workers have put horseshoes on the Keysian flea. School of Medicine, Vanderbilt University, Nashville, Tennessee 37232, USA
G. V. MANN
AGE-RELATED INCREASE IN AUTOANTIBODIES
SIR,-The apparent increase in frequency of autoantibodies with age was questioned by Dr Gordon and Dr Rosenthal (Jan 28, p 231). In a system testing anti-DNA antibodies and rheumatoid factor, they found comparable titres in healthy elderly subjects (aged over 70 years) and younger controls. If an increase in autoantibodies with age reflects the ageing process itself, it would be interesting to study autoantibodies in very old healthy people (aged over 90), elderly patients (aged under 90), and younger controls. Therefore we assessed antithyroglobulin antibodies (ATAb) (by means of passive haemagglutination) and antinuclear antibodies (ANAb) (with indirect immunofluoresence on mice cells) in three groups of subjects: healthy controls aged 20-50, elderly patients Coronary Drug Project Research Group-the CDP. Clofibrate and niacin in coronary heart disease. JAMA 1975, 231: 360-81. 2. Committee of the Principal Investigators, WHO Clofibrate Trial WHO cooperative 1.
trial
on
primary prevention of ischaemic heart disease using clofibrate
to
lower
cholesterol; mortality follow-up report. Lancet 1980; ii: 379-85 Lipid Research Clinics Program. The LRC coronary primary prevention trial results. JAMA 1984; 251: 351-64 Hyermann I, Velve Byre K, Holme I, et al. Effect of diet and smoking intervention on the incidence of CHD: report from the Oslo Study Group ofa randomized trialin healthy men. Lancet 1981, ii: 1303-10 Multiple Risk Factor Intervention Trial Research Group MRFIT. Risk factor changes and mortality results JAMA 1981, 248: 1465-77. serum
3. 4.
5.
Hypertension detection and follow-up program. Five year findings of the HDFP II Mortality by race-sex and age. JAMA 1979; 242: 2572-77 7 Grande F, Anderson JT, Chlouverakis C, Praja M, Keys A Effect of dietary cholesterol on man’s serum lipids. J Nutr 1965; 87: 52-62 8. Platt R. Ethics, responsibility and medical science. Australas Ann Med 1965; 14:
6.
93-95. 9. Massie S. Land of the firebirds New York. Simon and
Schuster, 1980