Upsala J Med Sci 8 1 : 85-92, 1976 Dietary and Drug Treatment of Hyperlipidaemia A Feasibility Study in Asymptomatic Middle-aged Men HANS HEDSTRAND From the Departments of Internal Medicine and Geriatrics, University Hospital, Uppsala, Sweden ABSTRACT 83 middle-aged men with different types of hyperlipopro- kinaemia were recruited from a health examination survey. They were treated with diet for 3 months and with diet and drugs in combination over a 2-year period (63 men used drugs). The serum lipid reductions after the dietary period were 14 % and 27 % for serum cholesterol and triglycerides, respectively. After 2 years the corresponding reductions were 21 % and 42 %, indicating an additional effect of diet and drugs. There was an average body weight reduction of 4.3% during the first 3 months which was maintained over the 2-year period. Special considerations in treating asymptomatic individuals are discussed. Prospective epidemiological studies have shown an association between elevated levels of cholesterol (8, 20) and triglycerides (TG) in serum ( 2 ) and de- velopment of coronary heart disease (CHD). Significant positive linear correlations between serum cholesterol and/or TG concentrations and risk of CHD have been demonstrated. During the last 20 years several feasibility studies (5, 11, 13) as well as primary preventive trials have been performed aiming at lowering serum choles- terol by dietary regimen (4, 12, 17, 19) or drug ther- apy (15, 16). The results as t o the lipid lowering ef- fects have vaned in different populations (14). The aim of the present study was to investigate the effect of dietary and drug treatment on serum lipid levels in a group of asymptomatic middle-aged men with hyperlipidaemia. MATERIAL All men, born 1920-1924, who lived in the city of Uppsala were invited to a health examination survey with the main purpose of identifying and correcting risk factors for car- diovascular disease (6). The investigation was conducted at the Department of Medicine, University Hospital. The participation rate was 83.9 %. A total number of 174 asymptomatic men with “pri- mary” hyperlipidaemia, defined as mean value of two samples of serum cholesterol and/or TG above the 80th percentile of the examined population, were invited to a special lipid clinic. One subject did not come t o the lipid clinic. Five men came to the first visit but did not come for the complete serum lipoprotein (LP) analysis. The result of the classification of the L P pattern in the remaining 168 subjects is shown in Table I. All 168 subjects were offered treatment after the L P analysis. Those with a normal L P pattern according to the limits used were given dietary prescriptions only and their serum lipids were reexamined after one year. This group comprised 30 men. All subjects with hyperlipoproteinae- mia (HLP) were initially treated with a dietary regimen for 3 months, after which they were offered additional therapy with lipid-lowering drugs. The results presented here concern the effect of the therapy on serum lipid concentrations in the group of 83 men with HLP repre- senting all men who a t the follow-up had participated for a t least 24 months after initiation of therapy. METHODS At the first visit to the lipid clinic the subjects were intro- duced to the programme and its long-term aspects were explained. The wives were invited together with the men to the second visit a t the lipid clinic. Results of the L P determinations were then presented and a dietary informa- tion was given by a dietitian, often to a group of three or four couples together. During the following three months of dietary treatment serum lipid analyses were performed at monthly intervals. The subjects were seen 2 weeks after each determination when they were informed of the result and had the possi- bility to see the dietitian if they desired. Diet prescriptions. The subjects were advised to avoid food high in ( a ) saturated fatty acids, e.g. fat meat and pork, dairy products such as butter, cream, cheese and whole milk and hard margarine. ( b ) cholesterol, e.g. egg yolks and liver. (c) simple sugars (mono- and disaccharides) e.g. refined sugar, sweets, sweet desserts, sweet beverages and sweet cakes (HLP type I1 B and IV). The subjects were advised to use foods containing ( a ) little or no saturated fatty acids, e.g. skimmed milk, lean meat, low-fat cheese. Upsala J M e d S c i 8 1 86 H . Hedstrand Table I . Lipoprotein patterns of 168 asymptomatic middle-aged m e n with serum lipids above the 80th percentile of the same population Lipoprotein No. of Subjects pattern subjects (%I Normal I1 A I1 B 111 IV V Total 37 22.0 38 22.6 25 14.9 15 8.9 51 30.4 2 1.2 168 100.0 ( b ) polyunsaturated fatty acids, e.g. vegetable oils (with more than 5 0 % linoleic acid), liquid margarine (containing more than 6 0 % of linoleic acid), soft margarine (contain- ing 2 5 4 5 % linoleic acid). The composition of the recommended diet should result in a P/S ratio of 2, total fat 30-35 % of the total calories and less than 300 mg cholesterol daily. The importance of using at least 50 ml of oil or liquid margarine daily was stressed. It was recommended to be used for frying, in dressings and in sauces. The subjects were advised that alcohol consumption should not exceed one bottle of beer a day. The daily caloric allowance was specified only when the subject had an actual over ideal weight > 1.20. Drug prescriprions. The subjects were offered drugs after the diet period, i.e. after 3 months. The drugs used were clofibrate (Atromidin@, ICI-Pharma, Goteborg, Sweden) and nicotinic acid (Nicangin@, Draco, Lund, Sweden). The men were informed about the most com- mon side effects of the drugs. Clofibrate was given in a dose of I g twice daily, as a rule in the morning and in the evening and not necessarily after meals. Treatment with nicotinic acid was started with 0.25 g three times daily after meals. The dosages were increased by further 0.25 g three times daily every fourth day until a total dose of 3 g daily was reached. The intention was to use clofibrate as the first drug in H L P type IV and nicotinic acid in H L P type I1 A and I1 B. Laborafory methods. At the screening examination the blood samples were drawn after an overnight fast. At the lipid clinic the blood samples were taken at the same time but the subjects were recommended to fast from 8 p.m. the evening before. Cholesterol and TG analyses were performed by semiautomatic techniques in a Technicon Auto-Analyzer type I1 (18). The separation and quantita- tive determination of L P density fractions by preparative ultracentrifugation as well as the analyses on agarose elec- trophoresis have been described in detail elsewhere (7). The L P patterns were classified according t o Fredrick- son et al. using the recommendations by Beaumont e t al. (1). The upper “normal” limits used, based on the 85th percentile of a local material of healthy men of different ages, were 200 mg/lOO ml for low density lipoprotein Ratio polyunsaturated fatty acids to saturated fatty acids. (LDL) cholesterol and 1.40 mmol/l for very low density lipoprotein (VLDL) TG (3). These limits were used by the lipid clinic and were close to those obtained in a sample of healthy men from the present population study (7). Sfatistical calculafions. Conventional methods were used for calculation of mean value and standard deviation (S.D.). Significances of differences between mean values were estimated with Student’s two-tailed t-test. When test- ing the means of serum TG concentrations the logarithm transformed values were used because the distribution of TG was skewed to the right (6). The differences between two values in the same subjects were estimated with the paired observation test using the Hewlett-Packard 9100 calculator programme. The accepted level of significance was p<0.05. RESULTS Subjects with normal LP pattern The screening values (mean of two samples) of serum cholesterol and TG for 30 subjects with nor- mal LP pattern at the lipid clinic were 299k33 mg1100 ml and 2.50k0.79 mmol/l, respectively. The average reduction between screening and the first analysis at the lipid clinic was 47 mg/100 ml(l4.9%) for serum cholesterol and 0.57 mmol/l (18.8%) for serum TG although no lipid lowering regimen had been prescribed. Table 11. Serum cholesterol concentrations ( m e a n M . E . M . ) in 83 asymptomatic middle-aged men with different types of hyperlipoproteinaemia ( H L P ) at initial examination and after 3 months of diet therapy and after 24 months of diet and drug therapy (drugs were used in 63 subjects) Months 0 3 24 H L P type I1 A ( n =20) Mean value, mg/lOO ml 3372 10 308f 11 2 6 0 5 9 Mean change, mg/100 ml - 2 9 f 2 . 3 7 5 f 2 . 5 Mean change, % - 8 . 6 f 3 . 0 2 1 . 3 f 3 . 0 H L P t y p e I I B (n=17) Mean value, mgll00 ml 3 2 1 f 8 2 7 7 2 7 2 4 4 2 8 Mean change, mg/100 ml - 4 4 f 2 . 1 7 7 f 2 . 4 Mean change, % - 1 3 . 2 f 2 . 4 23.5f2.8 H L P type III ( n = l O ) Mean value, mgll00 ml 3032 19 242k 17 2 2 9 f 9 Mean change, mg/100 ml - 6 2 k 5 . 9 7 5 f 4 . 1 Mean change, % - 1 9 . I f 5 . 6 2 3 . 4 f 2 . 8 H L P f y p e IV (n =36) Mean value, mg/lOO ml 2 7 8 f 6 2 3 5 f 7 2 2 1 f 6 Mean change, mg/lOO ml - 4 3 f 0 . 9 56k1.1 Mean change, % - 1 5 . 4 f 1 . 8 1 9 . 4 f 2 . 2 Upsala J M e d Sci 81 Dietary and drug treatment of hyperlipidaemia 87 Table 111. Serum triglyceride concentrations ( m e a n 5 S . E . M . ) in 83 asymptomatic middle-aged m e n with different types of hyperlipoproteinaemia ( H L P ) at initial examination and after 3 months of diet therapy and after 24 months of diet and drug therapy (drugs were used in 63 subjects) Months 0 3 24 H L P type I1 A (n =20) Mean value, mmol/l Mean value, (log) Mean change, mmol/l Mean change, % H L P t y p e I I B ( n = 1 7 ) Mean value, mmol/l Mean value, (log) Mean change, mrnol/l Mean change, % HLP type I l l ( n = I O ) Mean value, mmol/l Mean value, (log) Mean change, mmol/l Mean change, % H L P type IV ( n =36) Mean value, mmol/l Mean value, (log) Mean change, mmol/l Mean change, % 2.37 0.368k0.018 - - 3 .OO 0.471f0.019 - - 4.04 0.593k0.036 - - 3.36 0.513f0.018 - - 2.12 0.316f0.021 0.26k 0.03 9.7f4.7 2.18 0.326f0.025 0.82f 0.03 26.9f3.4 2.40 0.347f0.058 1.64k0.09 40.1f6.0 2. I9 0.316f0.024 1.18f0.02 33.7f 3.4 1.47 0.160f0.019 0.90f0.02 36.5k3.1 1.63 0.195f0.030 1.37f0.04 44.4k4.6 1.65 0.21 lf0.025 2.3950.12 56.4f4.5 1.85 0.246f 0.023 1.5 1 f 0 . 0 3 42.4f3.2 At the lipid clinic the subjects and their wives were given simple dietary recommendations by the physi- cian only. When these men were reexamined one year later they still showed significantly lower serum cholesterol and T G concentrations than at the initial screening. The average reductions were 31 mg/lOO ml(9.9 %) for cholesterol @<0.001) and 0.50 mmol/l (17.2%) for TG @<0.01) compared to the screening levels. After one year 90 % of the subjects had lower serum cholesterol and 77 % had lower TG than ini- tially. The mean body weight of the group was unchanged between the two visits at the lipid clinic one year apart. Subjects with hyperlipoproteinaemia Effect of dietary treatment. After 3 months of dietary treatment the serum cholesterol concentration was 262 mg/100 ml and the serum TG concentration was 2.19 mmol/l in the whole group of 83 men included in the 2-year study. These values corresponded t o a reduction of 13.7 % a n d 27.2 %for serum cholesterol and TG, respectively. The degree of serum lipid changes was depending on the type of serum LP pattern the patient had before the treatment started (Tables 11-111, Figs. 1-2). The greatest reductions, both in serum cholesterol and TG, were obtained in type 111 and the lowest in type I1 A . The changes of individual serum lipid values are presented in Figs. 4-7. A few subjects with type I1 A had a higher serum cholesterol value after three months of dietary therapy compared with the initial values. Ten men with type I1 A had unchanged or higher serum TG values. Among patients with type I1 B pattern only one had an increased serum choles- terol value and none had higher serum TG after three months. Two men with type 111 had slightly higher serum cholesterol values while all serum TG values were considerably reduced during the dietary period. All but 2 individuals with type IV had lower serum cholesterol and TG values after the three months of dietary treatment. The average body weight reduction was 4.3 % in the whole group during the first 3 months. The weight was reduced in all types of HLP (Table IV, Fig. 3). The reduction was greatest in subjects with type I1 B, 4.6 kg, and lowest in those with type I1 A, I .7 kg. E f f e c t of combined lipid lowering therapy. Sixty- Upsala J Med Sci 81 88 0 -20 -LO -60 H . Hedstrand 3 12 2LMonths ‘10 Change Fig. 1 . Effect of lipid lowering therapy on serum cholesterol over a 2-year period in 83 asymptomatic middle-aged men with type I1 A hyperlipoproteinaemia (n=20, 0-0), type I1 B (n=17, 0-O), type I11 (n=lO, A-A), and type IV (n=36, A-A). (Mean fS.E.M.) three of the 83 men (75.9%) who were followed over the 2-year period were treated with lipid-lowering agents. In 44 men (69.8%) drug therapy was started immediately after the dietary period, i.e. after 3 months. In 8 subjects drugs were introduced at 6 months and in another 8 men between 6 and 12 months. Only 3 men were given medication later than one year after the hypolipidaemia treatment was started. Table IV. Body weight (mean 3 . E . M . ) in 83 asymptomatic middle-aged men with different types of hyperlipoproteinaemia (HLP) a t initial examina- tion and after 3 months of diet therapy and after 24 months of diet and drug therapy (drugs were used in 63 subjects) Months 0 3 24 HLP t y p e I1 A ( n =20) Mean value, kg Mean change, kg Mean change, % HLP type I1 B (n=17) Mean value, kg Mean change, kg Mean change, % HLP t y p e I l l ( n = I O ) Mean value, kg Mean change, kg Mean change, % HLP type IV (n=36) Mean value, kg Mean change, kg Mean change, % ~ 73.25 2.1 - - 79.9f 3 . 3 - - 82.6k3.5 - - 81.3k 1.5 - - ~~ 71.5k2.0 1.7$0.1 2 . 2 f 0 . 5 75.3k 3.0 4 . 6 f 0 . 2 5 . 6 f 0 . 7 78.7f 3 . 3 3.9k0.2 4 . 7 f 0 . 7 77.7k 1.4 3.8f0.1 4 . 7 f 0 . 5 ~~ 70.8f 1.9 2.4k0.2 3.1k0.9 74.6k2.8 5.4k 0.3 6.4k 1.3 78.5k2.8 4.1k0.4 4.7k 1.4 76.2k 1.4 5 . l k 0 . 1 6.2k0.7 3 12 24 Months 0 -10 -20 I I -30 %Change F i g . 2. Effect of lipid lowering therapy on serum triglyce- rides over a 2-year period in 83 asymptomatic middle-aged men with hyperlipoproteinaemia. Symbols as in Fig. 1 . Nicotinic acid was given to 36 subjects. Their initial HLP types were type I1 25 cases, type 111 2 cases and type IV 9 cases. Clofibrate was given to 27 men, 7 with type I I , 4 with type I11 and 16 with type IV. In type I1 A and I1 B 32 men (86.5 %) used drugs. The percentage was lower, 69.4 %, in type IV where 25 subjects were given drug therapy. There were 9 men who declined drug therapy and 11 men were not offered drug therapy because the serum lipid concentrations on dietary treatment were already down to levels considered satisfactory. The mean serum lipid values after 3 months of dietary treatment in these 11 subjects were 197 mg/lOO ml and 1.66 mmol/l for cholesterol and TG, respectively. The LP pattern in 3 of these men was classified as type I11 and as type IV in 5. After 2 years the average serum cholesterol and TG concentrations were 236 mg/100 ml and 1.68 mmolll respectively in the whole group of 83 men. These reductions corresponded to 2 1.2 % and 43.1 %. The serum lipid reductions in the 63 men 3 12 24 Months I 1 -10 %Change Fig. 3. Reduction of body weight in 83 asymptomatic middle-aged men with hyperlipoproteinaemia during 2 years of lipid lowering therapy. Symbols as in Fig. 1 . Upsala J Med Sci 81 Dietary and drug treatment of hyperlipidaemia 89 who were on diet and drug therapy in combination were 22.9 % and 45.2 % for serum cholesterol and TG, respectively. The serum lipid changes in the four groups with different initial LP patterns are shown in Tables 11-111 and Figs. 1-2. The serum cholesterol reduc- tion was similar, around 75 mg/lOO ml, in all H L P types except for type IV. In all types the reduction was around 20% after 2 years. The greatest reduc- tion of serum TG was obtained in type 111, 2.39 mmol/l, which corresponded to 56.4 %. The individual changes of serum lipid concentra- tions after 2 years are shown in Figs. 4-7. In type I1 A (Fig. 4) 5 patients, all treated with drugs, had serum cholesterol values above the limit used as cut-off point (291 mg/100 ml) at the screening. The serum TG values were below the cut-off point (2.64 rnmol/l) in all patients. In type I1 B two individuals, both on drug therapy, had serum cholesterol values above the cut-off limit while the serum TG values were below the screening limit in all patients (Fig. 5). In a few subjects the serum lipids were higher than after the dietary period. All the subjects with type 111 had serum lipid values below the cut-off limits (Fig. 6). The serum cholesterol values of all men with type IV were below the cut-off values (Fig. 7). However, there were 14 men with higher values after 2 years than they had after the dietary period. Most of them had initially low serum cholesterol concentrations. Nine of them were without drug therapy. In 11 subjects the serum TG had increased after the first 3 months. Three individuals had values above the screening limit. The serum lipid values of the 11 men without drug therapy were 210+31 mg/lOO ml and 1.60k0.44 mmol/l for serum cholesterol and serum TG, respec- tively, corresponding to a reduction of 20.8% and 46.8 %. The 9 men who declined advice of drug therapy had significantly higher serum TG values after 2 years than the remaining subjects. The values were 2.20 mmol/l and 1.62 mmol/l, respectively ( p < O . O O I ) . The serum cholesterol values were not significantly different. The average body weight reduction in the total group after 2 years was 4.4 kg corresponding to 5.3 %. The reductions in the different types of H L P are shown in Table IV and Fig. 3 . The initial reduc- tions after 3 months were maintained throughout the CHOL TG 3 24 Months rnmo 1 - 1 I 3 24 I I Months F i g . 4 . Serum cholesterol and triglyceride concentrations in 20 men with type I1 A hyperlipoproteinaemia after 3 months of dietary therapy and after 2 years of diet and drugs in combination. O=diet, O=diet+clofibrate, A =diet+nicotinic acid, = diet +dofibrate +nicotinic acid C H O L rng / 1 OOml 1 T G rnrnc I : ',4 Months Fig. 5 . Serum cholesterol and triglyceride concentrations in 17 men with type I I B hyperlipoproteinaemia after 3 months of dietary therapy and after 2 years of diet and drugs in combination. Symbols as in Fig. 4. Upsulu J M e d Sci 81 90 H . Hedstrand CHOL r n g l l O O rnl 1 r 300 200 Months TG rnrnol 3 24 r I Months F i g . 6. Serum cholesterol and triglyceride concentrations in 10 men with type 111 hyperlipoproteinaemia after 3 months of dietary therapy and after 2 years of diet and drugs in combination. Symbols as in Fig. 4. 2 years. The greatest reduction, 5.4 kg, was obtained in subjects with type I I B and the lowest, 2 . 4 kg, in those with type I1 A . Out of the 168 men in whom the LP pattern was classified (Table I), 9 men (5.4%) dropped out during the following 18 months. If the six men who did not come for the L P analysis, are included the total drop-out rate over a 18-month period was 8.6% of those invited to the lipid clinic. Side effects of drug therapy. Nicotinic acid was discontinued in 19 men (52.8%) because of side effects. The reasons were: flushing in 13 cases (com- bined with pruritus in 3 cases) and gastrointestinal irritation in 5 cases. One withdrawal was caused by psychosomatic complaints, this patient complained of blurred vision. In 9 men nicotinic acid treatment was discon- tinued within 2 months, in 6 of them because of flushing. In the remaining 10 subjects the drug was withdrawn on average after 1 1 months (range 6-22). Liver function tests were checked regularly. In 3 cases there was a slight elevation of alkaline phos- phatase. No case of hyperuricaemia was seen. Clofibrate was well tolerated by all patients. When nicotinic acid had to be withdrawn, clofibrate was given instead. DISCUSSION The hyperlipidaemia found in this population was moderate. The L P pattern was classified as normal in 22 % of the subjects with lipid values over the 80th percentile at the initial examination. In 30 of them a lipid reduction of 15 % in serum cholesterol and 19 % in serum T G had occurred between the screening and the first visit to the lipid clinic, i.e. usually within 3 months. This reduction was obviously caused by dietary changes induced by the information of ele- vated serum lipids. A moderate but significant serum lipid reduction was maintained over a one-year period. This indicates that among subjects with hyperlipidaemia there is a number of individuals, perhaps 20 %, with moderate hyperlipidaemia in whom a substantial lipid reduction can be achieved by simple dietary recommendations. The actual serum cholesterol reduction of 10% in these 30 men can be compared to that of the National Diet Heart Study where the cholesterol level was reduced by 1 1 % over a one-year period (13). How- . ever, more impressive reductions have been re- ported in patients with CHD. In male survivors of myocardial infarction a serum cholesterol reduction of 17.6 % over a 5-year period was obtained with diet therapy alone (9). CHOL mg/100ml I 1 $ 24 0 I Months TG mmol/l 1 3 I 24 I Months F i g . 7 . Serum cholesterol and triglyceride concentrations in 36 men with type IV hyperlipoproteinaemia after 3 months of dietary therapy and after 2 years of diet and drugs in combination. Symbols as in Fig. 4. Upsala J Med Sci 81 There are several advantages, reviewed by Levy et al. (lo), in defining the type of H L P that is causing the hyperlipidaemia. Thus, it is well known that the different types of HLPdiffer in responsiveness t o the therapy. This was illustrated during the diet therapy by a serum cholesterol reduction of 19% in subjects with H L P type I11 compared to 9 9% in men with H L P type I I A . The serum TG reductions ranged from 10% in type I I A to 40% in type 111. The average serum lipid reductions obtained by diet therapy were maintained or increased over the 2-year period when drugs were added. Sixty-three men were given drug therapy. The effect of drugs may to some extent have been counteracted by decreased adherence to diet. On the other hand, the unchanged mean weight in all types of H L P may indicate that the instituted dietary regimen was fol- lowed satisfactorily. The reductions obtained in the present study by diet and drugs in combination (23 9% for serum cholesterol and 45% for TG) could be compared to those of 18 % for serum cholesterol and 28 '3% for serum TG reported from a primary preven- tive trial using clofibrate without advising any change in the usual diet ( 1 5 ) . Concerning drug therapy, most subjects accepted this therapy after the diet period. However, there were individuals who wanted to continue with diet alone for a longer period. No attempts were made to dissuade them. Later it was generally easier t o moti- vate them to accept drug therapy. There were also 9 men who for different reasons were not willing to use drugs. Nicotinic acid was discontinued due to side effects in over half of the subjects who had received this medication. Flushing was the most common side effect. Some subjects could not overcome the initial flushing. Others discontinued the drug because of flushing accentuated by irregularities in work and meals. This is an important aspect when a drug causes discomfort when not taken regularly, espe- cially when introducing medications in asympto- matic individuals. Clofibrate, on the other hand, was well tolerated. This drug has also the ad- vantage of being administered twice daily why the subjects do not have to take their medication during work. A high participation rate at the screening examina- tion and few drop-outs during the therapeutical trial indicates that middle-aged men are concerned about their health. A low frequency of side effects to the therapy is of particular importance when dealing Dietary and drug treatment ofhyperlipidaemia 91 with asymptomatic individuals, who may already have a tendency to blame the regimen, diet or drug, to cause every possible complaint. This study indicates that an acceptable degree of dietary adherence and patient cooperation can be accomplished in the treatment of asymptomatic middle-aged men with hyperlipidaemia. The serum lipid reductions can be maintained at least over a 2-year period. ACKNOWLEDGEMENTS This study was supported by Uppsala Lans Landsting, Pharmacia AB and grants from the Swedish Medical Re- search Council (19X-3116). REFERENCES 1 . Beaumont, J . L., Carlson, L. A,, Cooper, G. R., Fejfar, Z., Fredrickson, D. S . & Strasser, T.: Classifi- cations of hyperlipidaemias and hyperlipoproteinae- mias. Bull Wld Hlth Org43.891, 1970. 2. Carlson, L . A. & Bottiger, L . E.: Ischaemic heart 3. 4. 5 . 6 7. 8. 9. disease in relation to fasting values of plasma triglyc- erides and cholesterol. Lancet I : 865, 1972. Carlson, L. A. & Ericsson, M.: Quantitative and qual- itative serum lipoprotein analysis. I . Studies in healthy men and women. Atherosclerosis. In press, 1975. Dayton, S . , Pearce, M. L., Hashimoti, S . , Dixon, W. J . & Tomiyasu, U.: A controlled clinical trial of a diet high in unsaturated fat in preventing complications of atherosclerosis. Circulation 3 9 4 0 : Suppl. 2, 1969. Evans, D. W., Turner, S . M. & Ghosh, P.: Feasibility of long-term plasma-cholesterol reduction by diet. Lancet I : 172, 1972. Hedstrand, H.: A study of middle-aged men with particular reference to risk factors for cardiovascular disease. Upsala J Med Sci, Suppl. 19, 1975. Hedstrand, H . & Vessby, B.: Serum lipoprotein con- centration and composition in healthy 50-year old men. Upsala J Med Sci, in press. Kannel, W. B., Garcia, M. J., McNamara, P. M. & Pearson, G.: Serum lipid precursors of coronary heart disease. Human pathol2: 129, 1971. Leren, P.: The effect of plasma cholesterol lowering diet in male survivors of myocardial infarction. Acta Med Scand. SUDDI. 466. 1966. , _ _ 10. Levv. R. I . . Fredrickson. D. S . . Shulman. R.. Bil- 1 1 . 12. heimer, D. W.,Breslow, J. L . , Stone, N. J., Lux, S. E., Sloan, H. R., Krauss, R . M. & Herbert, P. N.: Dietary and drug treatment of primary hyperlipoproteinaemia. Ann Int Med 77: 267, 1972. McGandy, R. B., Hall, B., Ford, C. & Stare, F. J.: Dietary regulation of blood cholesterol in adolescent males: A pilot study. Amer J Clin Nutr 25: 61, 1972. Miettinen, M., Turpeinen, O., Karvonen, M. J., Elosuo, R. & Paavilainen, E.: Effect of cholesteroI- lowering diet on mortality from coronary heart disease and other causes. A twelve-year clinical trial in men and women. Lancetll: 835, 1972. Upsala J M e d Sci 81 92 H . Hedstrand 13. National Diet Heart Study. Final report. Circulation 37: Suppl. I, 1968. 14. National Heart Foundation of Australia. Dietary fat and coronary heart disease: A review. Med J Australia 58: 1155, 1971. 15. Oliver, M. F.: A primary prevention trial using clofibrate t o lower hyperlipidaemia. I n Atheroscle- rosis: Proceedings of the Second International Symposium (ed. R. J. Jones). Springer-Verlag, Berlin, Heidelberg and New York, 1970. 16. Rifkind, B.: Drugs affecting lipid metabolism. Milano, Sept. 1974. 17. Rinzler, S. H.: Primary prevention of coronary heart disease by diet. Bull NY Acad Med44: 936, 1%8. 18. Rush, R. L., Leon, L. & Turrell, J.: Automated simultaneous cholesterol and triglyceride determina- tion on the Auto-Analyzer I1 Instrument. Advances in automated analysis. Thurman associates I : 503, 1971. 19. Stamler, J.: Acute myocardial infarction-progress in primary prevention. Brit Heart J33, Suppl. 145, 1971. 20. Westlund, K . & Nicolaysen, R.: Ten-year mortality and morbidity related to serum cholesterol. A follow- up of 3 751 men aged 4 0 4 9 . Scand J Clin Lab Invest, Suppl. 127, 1972. Recieved J a n u a r y 5, 1976 Address for reprints: Hans Hedstrand, M.D. Department of International Medicine University Hospital S-750 14 Uppsala Sweden Upsala J M e d Sci 81