Postgrad Med J (1991) 67, 931 - 937 i The Fellowship of Postgraduate Medicine, 1991

Treatment of diabetes: the effect on serum lipids andlipoproteins

Peter K. Merrin and Robert S. Elkeles

Department ofMetabolic Medicine, St. Mary's Hospital, Praed Street, London W2 INY, UK

Introduction

In both insulin dependent diabetes (IDDM) andnon-insulin dependent diabetes (NIDDM), mor-bidity and mortality from cardiovascular disease isgreatly increased. The causes of the acceleratedcardiovascular disease are probably different in thetwo types. As yet there is no evidence that controlof hyperglycaemia has reduced the incidence ofcardiovascular disease in either type.' There is alarge body of evidence that hyperlipidaemia in thegeneral population is associated with cardiovas-cular disease. Furthermore there is considerableevidence that control of serum lipids results in thereduction of the incidence of coronary heartdisease2'3 and that in those with established coron-ary heart disease, the rate of progression is reduc-ed.4 It is therefore important to understand theeffects of the treatments used in diabetes on serumlipids and lipoproteins.

Non-insulin dependent diabetes

Prevalence ofserum lipoprotein abnormalities

In NIDDM there is a two- to three-fold increase incardiovascular mortality' which accounts forbetween 50 and 75% of deaths.6 Risk factors forcardiovascular disease in the general population,such as hypertension, smoking and raised serumcholesterol, appear important in NIDDM.5 Cross-sectional studies have indicated that increasedserum triglyceride, total and low density lipo-protein (LDL) cholesterol, low high densitylipoprotein (HDL) cholesterol and apoprotein Alare associated with cardiovascular disease inNIDDM.7'-0 Data from prospective studies haveshown that serum triglyceride is an important riskfactor.""2 In women, low HDL cholesterol,'3 andin middle aged men, serum cholesterol,'4 haveemerged as important predictors of cardiovascular

disease. The most common metabolic abnormali-ties in NIDDM are hypertriglyceridaemia and lowHDL cholesterol.'5 There are sex differences in thedistribution of the lipoprotein abnormalities inNIDDM.'6 The increase in serum triglyceride isrelatively greater and the decrease in HDL choles-terol relatively lower in women than in men.'6Hypertriglyceridaemia and low HDL cholesterol,often referred to as dyslipidaemia, are metabol-ically closely related and may well be important inthe aetiology of macrovascular disease in NIDDM(see also Winocour, this issue, pp. 917-921).

Diet

In the overweight, weight reduction reduces serumtriglycerides and raises HDL cholesterol.'7"18 In thenon-obese, diet may lower serum triglyceride buthas little effect on HDL cholesterol.'7"8 Currentadvice is to encourage a high fibre, high carbohy-drate, low saturated fat diet. It has been shown thata simple increase in digestible carbohydrate with-out a parallel increase in dietary fibre does not helpin improving metabolic control.'9 Indeed, somehigh carbohydrate diets have been shown to in-crease serum triglyceride, lower HDL cholesteroland worsen glucose control.20 The hypoglycaemicand hypolipidaemic effects of high carbohydrate,high fibre diets are due mainly to the fibre content.

It is important to remember that there are twotypes of fibre, soluble or viscous fibre and insolubleor fibrous fibre, such as bran. It is only soluble fibre,such as that contained in legumes and greenvegetables, which has the desired metabolic effectswhereas insoluble fibre has barely any. The phar-macological addition of fibre in the form of guarhas been shown to reduce total and LDL choles-terol,21'22 but the quantity of fibre recommended insome diets is not acceptable to many patients. Thebenefits and disadvantages of fibre in the manage-ment of diabetes have been recently well debat-ed.2324 Controversy therefore exists as to whichconstituent should replace the reduced saturatedfat intake which is considered desirable. There issome evidence that partial replacement ofcomplexCorrespondence: R.S. Elkeles, M.D., F.R.C.P.

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from

932 P.K. MERRIN & R.S. ELKELES

carbohydrate with mono-unsaturated fat improvesglucose control, improves serum triglyceride andHDL cholesterol without raising LDL choles-terol.25 The current enthusiasm for high carbohy-drate diets inNIDDM may need slight modification.Longer term studies are needed to define the mostsuitable diet for people with NIDDM.

Oral hypoglycaemic drugs

Doubts were expressed about the long term safetyof the sulphonylureas when the American Univer-sity Group Diabetes Program Study showed that inpeople with NIDDM who were randomized toreceive either diet alone, diet and tolbutamide ordiet and insulin, the tolbutamide group had agreatly increased cardiovascular mortality com-pared with the other groups.26 The findings of thisstudy have been disputed.27'28 Nevertheless, theeffects of sulphonylureas on serum lipids wereinvestigated. Early cross-sectional studies showedthat those taking sulphonylureas had lower HDLcholesterol than those on diet or insulin.29'30Another study showed that those on sulphonyl-ureas had both higher triglycerides and lower HDLcholesterol than those on diet or insulin.3' How-ever, subsequent prospective studies have failed toconfirm these findings. Neutral or even beneficialeffects of sulphonylureas have been found. Paiseyet al.32 found a reduction in serum and very lowdensity lipoprotein (VLDL) triglyceride with a risein HDL cholesterol with chlorpropamide, whileGreenfield et al.33 found a reduction in serumtriglyceride with glipizide, but no change in HDLcholesterol. Taylor et al.'8 found no significantchange in serum lipids with glibenclamide thoughthere was a rise in apoprotein A1. Rains et al.,34comparing glibenclamide with metformin, foundno significant change in serum lipids and lipopro-teins with glibenclamide. It is of interest thatGlyburide (glibenclamide) given to insulin treatednon-insulin dependent diabetics35 had no signi-ficant effects on lipids and lipoproteins. Taking allthe evidence together, sulphonylureas probablyhave little overall effect on serum lipids and do notcorrect the low HDL cholesterol of NIDDM.37

Metformin is the only biguanide drug in com-mon use in the United Kingdom. It has been shownto have a small effect in reducing total and LDLcholesterol.34 Furthermore, this effect was found tobe maintained in the long term.38 In another studymetformin was found to reduce onlyVLDL choles-terol in NIDDM.2' However, in those with acholesterol greater than 6.5 mmol/l, metforminsignificantly reduced serum triglyceride.22 Wu et al.have shown that metformin lowered both fastingand post-prandial serum triglyceride and evenincreased HDL cholesterol in a group of patientswith NIDDM.39 In addition metformin has been

shown to reduce serum triglyceride in non-diabeticsubjects.'" These data therefore suggest that met-formin does have a small effect in improving thelipid profile in patients with NIDDM.

Insulin

The decision as to when to use insulin in patientswith NIDDM can be difficult. Insulin has beenshown to reduce serum triglyceride and cholesterol,but not affect HDL cholesterol in a small group ofpatients with NIDDM.4' Lindstrom et al.42 com-pared two different insulin regimes in a group ofpatients with NIDDM who were secondary sul-phonylurea failures. Insulin was given as either a 2or 4 dose regime. Both insulin regimes resulted insimilar improvements in blood glucose control.They also resulted in a decrease in serum choles-terol, triglyceride and increases in total HDL,HDL2 and HDL3 cholesterol as well as apolipopro-tein Al. Insulin therapy has its disadvantages,particularly in the obese subject and, clearly itcannot be used in the majority of patients withNIDDM.

Lipid-modifying drugs

There remain a large number of subjects withNIDDM who have a modestly increased serumtriglyceride and reduced HDL cholesterol, despitediet and oral hypoglycaemic drugs. The importantquestion, therefore, arises as to whether suchpatients should be given a lipid-modifying agent. Inthe Helsinki Heart Study,3 middle aged men withmild hyperlipidaemia similar to that seen inNIDDM were randomized into those receivingplacebo or the fibrate gemfibrozil. There was asignificant reduction in coronary events in theactively treated group during the 5-year period offollow-up. Gemfibrozil therapy was associatedwith a persistent increase in HDL cholesterol andreductions in serum total and LDL cholesterol andin triglyceride. Another fibrate, bezafibrate hasbeen shown to reduce serum triglyceride andcholesterol and increase HDL cholesterol inNIDDM.8 One of the new class of HMG CoAreductase inhibitors, lovastatin, has been shown toreduce total and LDL cholesterol and LDL apo-protein B and triglycerides in NIDDM.43 The effecton HDL cholesterol was variable. A recent studycompared the effects of gemfibrozil and lovastatinon lipids and glucose control in NIDDM.4 Lovas-tatin significantly reduced total and LDL choles-terol and raised HDL cholesterol but did notreduce triglyceride, while gemfibrozil did not signi-ficantly reduced LDL cholesterol. Neither drugaffected blood glucose control. It is difficult topredict which class ofdrug will be the most effectivein this situation. However, since the dyslipidaemia

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from

DIABETES: TREATMENT AND SERUM LIPIDS 933

ofNIDDM is mainly elevated triglyceride and lowHDL cholesterol, the fibrate group of drugs wouldseem to be more appropriate. There is a need forlong term prospective studies in NIDDM to ascer-tain whether, by improving the dyslipidaemia withdrug therapy, the progress of cardiovasculardisease can be slowed and its incidence reduced.

Insulin dependent diabetes

Prevalence ofserum lipid abnormalities andfactorsinfluencing it

In the pre-insulin era, death for those with juvenileonset diabetes was invariably due to ketoacidosis.However, as insulin treatment became available,there was a large increase in the numbers ofdiabetics dying from cardiovascular disease.5 Thiswas even more evident in the female diabeticpopulation.45 Much of this increase in cardiovas-cular mortality is from coronary artery disease.46-4'There is a paucity of information on the relation-ship between serum lipids and cardiovasculardisease in IDDM. Furthermore, there needs to be aclear distinction drawn between the insulin treatednon-insulin dependent and insulin dependentpopulation. However, studies have indicated thatrisk factors for coronary artery disease in theinsulin dependent diabetic include elevated LDLcholesterol,9 antecedent nephropathy49 and hyper-tension.' It is of interest that no inverse relation-ship between HDL cholesterol concentrations andvascular disease was found in the male insulintreated diabetic.9 An early study5" indicated thatmiddle-aged diabetic patients treated with insulinfor a minimum period of 10 years had serumcholesterol and triglyceride concentrations similarto that of non-diabetic controls. However, theconcentration ofHDL cholesterol was significantlyhigher in male and female diabetics than therespective control groups.The prevalence of hyperlipidaemia in 205 insu-

lin-dependent individuals was examined by Wino-cour et al.52 They found that 40% ofthe individualstested had some type ofhyperlipidaemia. While theprevalence of hypertriglyceridaemia and combinedhyperlipidaemia was greater in patients with insu-lin-dependent diabetes than non-diabetics, this wasnot the case for hypercholesterolaemia. In thisstudy, hypertriglyceridaemia was associated with adecrease in the HDL2 cholesterol compared topatients with normal serum lipids or hypercholes-terolaemia. Nikkila53 demonstrated low HDL con-centrations in untreated type I diabetics, but mostother studies in treated patients have demonstratednormal or above normal levels of this lipopro-tein. 54-6 The development of nephropathy has anadverse effect on serum lipids. Patients with albu-

minuria have higher total and LDL cholesterol,higher triglycerides and apolipoprotein B levelstogether with lower HDL cholesterol concentra-tions compared to those without albuminuria.57Similar findings were reported in a group ofpatients with microalbuminuria.58 The presence ofproteinuria in IDDM is associated with an in-creased mortality.59 In an analysis of 2,890 patientswith proteinuria,W the relative mortality fromcardiovascular disease was 37 times that of thegeneral population. In those without proteinuria itwas 4.2 times that in the general population. Theinfluence of proteinuria on vascular disease, bloodpressure and lipoproteins in IDDM was studied byWinocour et al.6' Of 22 patients who had pro-teinuria (but not renal failure), 10 had evidence ofmacrovascular disease, whilst this was only presentin three patients with normal albumin excretionrates. The patients with proteinuria had higherblood pressures, lower serum HDL concentrations(as measured by a precipitation technique), lowerHDL2 concentrations (as isolated by ultracentri-fugation) and high HDL3 concentrations.

Insulin therapy

Lipid and lipoprotein abnormalities in insulindependent diabetes depend on the degree of insu-linization of the individual. In the totally insulindeficient state, i.e. ketoacidosis, gross hypertrigly-ceridaemia with greatly elevated levels of VLDLcholesterol can occur.62'63 Weidman et al. showed,in a group of 13 patients, that these abnormalitiescan largely be corrected within the 24 hours afterthe initiation of insulin treatment.' Mean levels ofintermediate density lipoprotein (IDL) cholesteroland LDL cholesterol were low and did not varywith therapy. Mean basal levels of HDL choles-terol were also low and did not vary during the first12 hours ofinsulin therapy. However, from 12 to 24hours the levels rose significantly. Plasma apolipo-protein B levels fell with the therapy due tosignificant decreases in VLDL cholesterol.

Intensive insulin therapy

Intensive insulin therapy was found to improvelipid profiles by reducing the concentrations oftotal, LDL and VLDL cholesterol.65 The effects onHDL cholesterol in this study were not as pro-nounced, although there were rises, particularly inmale patients with poor overall control. In acomparison of multiple injection therapy and con-tinuous subcutaneous insulin infusion (CSII) byPietri et al.66 both treatment modalities reducedserum triglyceride levels but only CSII reducedserum and LDL cholesterol. Longer periods ofsubcutaneous insulin infusion are additionallyassociated with elevations in HDL cholesterol67,6

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from

934 P.K. MERRIN & R.S. ELKELES

and a reduction in plasma levels of apolipoproteinB.77 In a study of 10 normolipidaemic patientsbefore and after improvement in metabolic con-trol,69 changes in VLDL and LDL cholesterolcomposition were noted. Subfractions from theIDDM patients were cholesteryl-ester poor andtriglyceride rich (particularly in the LDL subfrac-tions) and contained higher proportions of non-apolipoprotein B apoproteins compared to thecontrol group. The apolipoprotein abnormalityresponded to improved control, but did not norma-lize, and, although LDL and VLDL concentrationsimproved, the compositional abnormalities per-sisted.

Diet

In a study of 270 children by Kaufmann et al.70 theeffects of a usual and modified diabetic diet onglycaemic control and serum lipids was assessed.The usual diabetic diet contained 700-1500 mg ofcholesterol daily with a polyunsaturated/saturated(P/S) fatty acid ratio of 0.1, while the modifieddiet limited cholesterol to 300 mg daily with aP/S ratio of 1.0. Both diets contained similaramounts of protein and carbohydrate as calories.On admission to the study 24% of the subjects hadhyperlipoproteinaemia (mostly combined hyper-lipidaemia or hypertriglyceridaemia). After follow-ing the usual diet, hypertriglyceridaemia was onlypresent in 1% of individuals, whereas 21% of thesubjects still had a type II electrophoretic pattern,despite overall lower glucose levels. After con-sumption of the modified diet, hyperlipoprotein-aemia was reduced to 5% with only 4% of subjectshaving a type II pattern and 1% hypertriglyceri-daemia.

Hollenbeck et al. studied the effects of diet inIDDM.7' They showed that after 6 weeks of a dietcontaining 65% of the calories as carbohydrate,plasma triglycerides were significantly increased,although this may not apply in well controlledpatients.72 Decreases in concentrations of apoli-poproteins Al, B and CIII were also shown in agroup of IDDM patients treated with a low fat,high carbohydrate, cholesterol restricted diet.7'There have been few studies on the prolongedeffects of dietary fibre on serum lipids in IDDM. Inone study of 8 insulin-treated patients who received14-26 grammes of guar/day there was a 15%reduction in LDL cholesterol without change in theserum HDL cholesterol." This was accompaniedby a mean reduction in the insulin dose required tomaintain glycaemic control. General considera-tions concerning the diet for IDDM, including thefibre content, are similar to those for NIDDM (seeabove).

Glycaemic control andserum lipidsA number of investigators have examined therelationship between glycaemic control, as mani-fested by glycosylated haemoglobin levels, andserum lipids. Improved glycaemic control reducesserum cholesterol and triglycerides.74'75 There doesnot appear to be a correlation between haemo-globin Al and HDL cholesterol,757" althoughLopes-Virella et al.78 found an inverse relationshipbetween haemoglobin AIc and HDL cholesterol inmen and women with poor overall diabetic control(as defined by haemoglobin Alc and 24 hourglycosuria). In a study of 544 patients haemoglobinAl was negatively correlated with HDL cholesteroland positively correlated with the levels of totalcholesterol, total triglycerides and LDL choles-terol.79 These relationships were not present inblack women suggesting that race and gender maybe independent determinants of serum lipids ininsulin-dependent diabetics.

Lipid modifying drugsThere have been few studies on the action ofhypolipidaemic agents in insulin-dependent dia-betes. Fibric acid derivatives have been advocatedas the drug of choice for diabetic dyslipidaemia.80In Type 1 diabetes, bezafibrate therapy caused asignificant reduction in fasting blood glucose,VLDL cholesterol, serum triglycerides and apoli-poprotein B.8' There was no overall change inHDLcholesterol, but in those patients who were initiallyhypertriglyceridaemic, there was a significant in-crease in the HDL2 subfraction. The use of fish oilcontaining omega 3 fatty acids has been reviewedrecently82 and there is some evidence for a beneficialeffect on vascular permeability,83 suggesting apossible prophylactic use of these agents in theprevention of atherogenesis. However, the resultsin other studies have not shown consistent benefitsin lipid profiles. In a group of male patients84therapy caused elevations of total cholesterol andapolipoprotein B. In a further study of femalepatients, a fall in plasma triglycerides and a risein HDL cholesterol was noted.85 Additionally, theuse of 3-omega fatty acids may be associated witha deterioration in glycaemic control.86 Furtherstudies on these and other newer hypolipidaemicagents in insulin-dependent diabetes are needed.The aetiology of macrovascular disease in IDDMappears more complex than in NIDDM and itsrelationship to dyslipidaemia is less well defined,although this may not be the case for those IDDMpatients with proteinuria. The use of lipid-modify-ing drugs in both NIDDM and IDDM is consider-ed in detail in another article in this series by PaulDurrington (this issue, pp. 947-952).

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from

DIABETES: TREATMENT AND SERUM LIPIDS 935

References

1. Pyorala, K. & Laakso, M. Macrovascular disease in diabetesmellitus. In: Mann, JI., Pyorala, K. & Teuscher, A. (eds),Diabetes in Epidemiological Perspective. Churchili Living-stone: Edinburgh, 1983, pp. 183-247.

2. The Lipid Research Clinic Program. The Lipid ResearchClinics Coronary Primary Prevention Study Trial Results II.The relationship of reduction in incidence of coronary heartdisease to cholesterol lowering. JAMA 1984, 251: 365-374.

3. Frick, M.H., Elo, O., Happa, K. et al. Helsinki Heart Study:primary prevention trial with Gemfibrozil in middle-agedmen with dyslipidaemia. NEnglJMed 1987, 317:1237-1245.

4. Blankenhorn, D.H., Nessim, S.A., Johnson, R.L., Sanmarco,M.E., Azen, S.P. & Cashin-Hemphill, L. Beneficial effects ofcombined Colestipol-Niacin therapy on coronary athero-sclerosis and coronary venous bypass grafts. JAMA 1987,257: 3233-3240.

5. Krannel, W.B. & McGee, D.L. Diabetes and cardiovasculardisease. The Framingham Study. JAMA 1979, 241: 2035-2038.

6. Panzram, G. Mortality and survival in Type 2 (non-insulindependent) diabetes. Diabetologia 1987, 30: 123-131.

7. Santen, R.J., Willis, P.W., Fajans, S.S. & Arbor, A. Athero-sclerosis in diabetes mellitus. Correlations with serum lipidlevels, adiposity and serum insulin level. Arch Intern Med1972, 130: 833-843.

8. Seviour, P.W., Teal, T.K., Richmond, W.R. & Elkeles, R.S.Serum lipids and macrovascular disease in non-insulin depen-dent diabetics: A possible new approach to prevention.Diabetic Med 1988, 5: 166-171.

9. Reckless, J.P.D., Betteridge, D.J., Wu, P., Payne, B. &Galton, D.J. High density and low density lipoproteins andprevalence of vascular disease in diabetes mellitus. Br Med J1978, 1: 883-886.

10. Laakso, M., Ronnemaa, T., Pyorala, K., Kallio, V., Puukka,P. & Penttila, I. Atherosclerotic vascular disease and its riskfactors in non-insulin dependent diabetic and non-diabeticsubjects in Finland. Diabetes Care 1988, 11: 449-463.

11. Janka, H.U. Five year incidence of major macrovascularcomplications in diabetes mellitus. Horm Metab Res 1985, 15(Suppl): 15-19.

12. Fontbonne, A., Eschwege, E., Cemblen, F. et al. Hypertrigly-ceridaemia as a risk factor for coronary heart diseasemortality in patients with impaired glucose tolerance ordiabetes. Results from the eleven year follow up of the ParisProspective Study. Diabetologia 1989, 32: 300-304.

13. Gordon, E., Casteli, W.P., Hjortland, M.C., Kannel, W.B. &Dawber, T.R. Diabetes, blood lipids, and the role of obesityin coronary heart disease risk for women. Arch Intern Med1977, 137: 393-397.

14. Rosengren, A., Welin, L., Tsipogianni, A. & Wilhelmsen, L.Impact of cardiovascular risk factors on coronary heartdisease and mortality among middle aged diabetic men: ageneral population study. Br Med J 1989, 299: 1127-1131.

15. Reaven, G.M. Non-insulin dependent diabetes mellitus,abnormal lipoprotein metabolism and atherosclerosis. Meta-bolism 1987, 36 (Suppl 1): 1-8.

16. Walden, C.E., Knopp, R.H., Wahl, P.W., Beach, K.W. &Strandness, E. Sex differences in the effect ofdiabetes mellituson lipoprotein triglyceride and cholesterol concentrations. NEngl J Med 1984, 311: 953-959.

17. Kennedy, L., Walshe, K., Hadden, D.R., Weaver, J.A. &Buchanan, K.D. The effect of intensive dietary therapy onserum high density lipoprotein cholesterol in patients withtype 2 (non-insulin dependent) diabetes mellitus, A prospec-tive study. Diabetologia 1982, 23: 24-27.

18. Taylor, K.G., John, W.G., Matthews, K.A. & Wright, A.D.A prospective study of 12 months treatment on serum lipidsand apolipoproteins A-1 and B in type 2 (non-insulindependent) diabetes. Diabetologia 1982, 23: 507-510.

19. Riccardi, G., Rivellese, A., Pacioni, D., Genovese, S., Mas-tranzo, P. & Mancini, M. Separate influence of dietarycarbohydrate and fibre on the metabolic control in diabetes.Diabetologia 1984, 26: 116-121.

20. Coulston, A.M., Hollenbeck, C.H., Swislocki, A.L.M. &Reaven, G.M. Deleterious metabolic effects of high carbo-hydrate, sucrose containing diets in patients with non-insulindependent diabetes mellitus. Am J Med 1987, 82: 213-220.

21. Lalor, B.C., Bhatnager, D., Winocour, P.H. et al. Placebo-controlled trial of the effects on guar gum and Metformin onfasting blood glucose and serum lipids in obese, type 2diabetic patients. Diabetic Med 1990, 7: 242-245.

22. Fuessl, H.S., Williams, G., Adrian, T.E. & Bloom, S.R. Guarsprinkled on food: effect on glycaemic control, plasma lipidsand gut hormones in non-insulin dependent diabetic patients.Diabetic Med 1987, 4: 463-468.

23. Hockaday, T.D.R. Controversies in Therapeutics. Fibre inmanagement of diabetes. 1. Natural fibre useful as a part oftotal dietary prescription. Br Med J 1990, 300: 1334-1336.

24. Tattersall, R. & Mansell, P. Controversies in Therapeutics.Fibre in the management ofdiabetes. 2. Benefits of fibre itselfare uncertain. Br Med J 1990, 300: 1336-1337.

25. Garg, A., Bonanome, A., Grundy, S., Zhang, Z. & Ungar,R.H. Comparison of a high carbohydrate diet with a highmono-unsaturated fat diet in patients with non-insulindependent diabetes mellitus. N Engl J Med 1988, 319:629-634.

26. University Group Diabetes Program. A study of the effects ofhypoglycaemic agents on vascular complications in patientswith adult-onset diabetes. Diabetes 1970, 19 (Suppl 2).

27. Kilo, C., Miller P & Williamson, J.R The crux of the UGDP:spurious results and biologically inappropriate data analysis.Diabetologia 1980, 18: 179-185.

28. Fuller, J.H. Clinical trials in diabetes mellitus. In: Mann, J.I.et al. (eds) Diabetes in Epidemiological Perspective. ChurchillLivingstone: Edinburgh, 1983, pp. 265-285.

29. Berry, E.M. & Bar-On, H. Comparison of sulphonylurea andinsulin treatment on lipid levels in maturity-onset diabeticmen and women. Isr J Med Sci 1981, 17: 384-387.

30. Calvert, G.D., Graham, J.J., Mannik, T., Wise, P.H. &Yeates, R.A. Effects of therapy on plasma- high-density-lipoprotein-cholesterol concentration in diabetes mellitus.Lancet 1978, if: 66-68.

31. Lisch, H.-J. & Sailer, S. Lipoprotein patterns in diet, sul-phonylurea and insulin treated diabetics. Diabetologia 1981,20: 118-122.

32. Paisey, R., Elkeles, R.S., Hambley, J. & Magill, P. The effects ofchlorpropamide and insulin on serum lipids, lipoproteins andfractional triglyceride removal. Diabetologia 1978, 15: 81-85.

33. Greenfield, M.S., Doberne, L., Rosenthal, M., Vreman, H.J.& Reaven, G.M. Lipid metabolism in non-insulin diabetesmellitus. Effect of glipizide therapy. Arch Intern Med 1982,142: 1498-1500.

34. Rains, S.G.H., Wilson, G.A., Richmond, W. & Elkeles, R.S.The effect of glibenclamide and metformin on serum lipopro-teins in type 2 diabetes. Diabetic Med 1988, 5: 653-658.

35. Falko, J.M. & Ossei, K. Combination insulin/Glyburidetherapy in type II diabetes mellitus. Am J Med 1985, 79(Suppl 3B): 92- 101.

36. Seviour, P.W., Teal, T.K., Richmond, W. & Elkeles, R.S.Chlorpropamide lowers serum and lipoprotein cholesterol ininsulin-dependent diabetes. Diabetic Med 1986, 3: 152-154.

37. Baynes, C., Feher, M. & Elkeles, R.S. The effect of treatmentof non-insulin dependent diabetes mellitus on serum lipidsand lipoproteins. Q J Med 1989, 72: 579-587.

38. Rains, S.G.H., Wilson, G.A., Richmond, W. & Elkeles, R.S.The reduction of low density lipoprotein cholesterol bymetformin is maintained with long-term therapy. J R SocMed 1989, 82: 93-94.

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from

936 P.K. MERRIN & R.S. ELKELES

39. Wu, M.-S., Johnston, P., Sheu, W.H.-H. et al. Effect ofmetformin on carbohydrate and lipoprotein metabolism inNIDDM patients. Diabetes Care 1990, 13: 1-8.

40. Sirtori, C.R., Tremoli, E., Sirtori, M., Conti, F. & Paoletti, R.Treatment of hypertriglyceridaemia with metformin. Effec-tiveness and analysis of results. Atherosclerosis 1977, 26:583-592.

41. Hollenbeck, C.H., Chen, Y.I., Greenfield, M.S., Lardinois,C.K. & Reaven, G.M. Reduced high density lipoproteincholesterol concentrations need not increase when hyper-glycaemia is controlled with insulin in non-insulin dependentdiabetes mellitus. J Clin Endocrinol Metab 1986, 62: 605-608.

42. Lindstrom, T., Arnqvist, H.J. & Olsson, A.G. Effect ofdifferent insulin regimes on plasma lipoprotein and apolipo-protein concentrations in patients with non-insulin depen-dent diabetes mellitus. Atherosclerosis 1990, 81: 137-144.

43. Garg, A. & Grundy, S. Lovastatin for lowering cholesterollevels in non-insulin dependent diabetes mellitus. N Engl JMed 1988, 318: 81-86.

44. Goldberg, R., La Belle, P., Zupkis, R. & Ronca, P. Com-parison of the effects ofLovastatin and Gemfibrozil on lipidsand glucose control in non-insulin dependent diabetes melli-tus. Am J Cardiol 1990, 66: 16B-21B.

45. Entmacher, P.S., Rout, H.F. & Marks, H.H. Longevity ofdiabetic patients in recent years. Diabetes 1964, 13: 373-377.

46. Jarrett, R.J. Risk factors ofmacrovascular disease in diabetesmellitus. Horm Metab Res 1985, 15 (Suppl): 1-3.

47. Dupree, E.A. & Meyer, M.B. Role of risk factors incomplications ofdiabetes mellitus. Am JEpidemiol 1980, 112:100-112.

48. Krolewski, A.S., Kosinski, E.J., Warram, J.H. et al. Mag-nitude and determinants of coronary artery disease injuvenile-onset, insulin-dependent diabetes mellitus. Am JCardiol 1987, 59: 750-755.

49. Jensen, T., Borch-Johnsen, K., Kofoed-Enevoldsen, A. &Deckert, T. Coronary heart disease in young type 1 (insulin-dependent) diabetic patients with and without diabeticnephropathy: incidence and risk factors. Diabetologia 1987,30: 144-148.

50. Stamler, J., Wentworth, D., Neaton, J., Schoenberger, J.A. &Feigal, D. for the MRFIT Research Group. Diabetes and riskofcoronary, cardiovascular and all causes mortality: findingsfor 356,000 men screened by the Multiple Risk FactorIntervention Trial (MRFIT). Circulation 1984, 70 (Suppl 2):161.

51. Nikkila, E.A. & Hormila, P. Serum lipids and lipoproteins ininsulin-treated diabetes. Demonstration of increased highdensity lipoprotein concentrations. Diabetes 1978, 27:1078- 1086.

52. Winocour, P.H., Durrington, P.N., Ishola, M., Hillier, V.F.& Anderson, D.C. The prevalence of hyperlipidaemia andrelated clinical features in insulin-dependent diabetes mel-litus. Q J Med NS 1989, 70: 265-276.

53. Nikkila, E.A. High density lipoproteins in diabetes. Diabetes1981, 30 (Suppl 2): 82-87.

54. Laakso, M., Voutilainen, E., Aro, A., Pyorala, K. & Penttila,I. Inverse relationship of serum HDL and HDL2 cholesterolto C-peptide level in middle aged insulin treated diabetics.Metabolism 1985, 34: 715-720.

55. Eckel, R.H., Albers, J.J., Cheung, M.C., Wahl, P.W., Lind-gren, F.T. & Bierman, E.L. High density lipoprotein com-position in insulin-dependent diabetes mellitus. Diabetes1981, 30: 132-138.

56. Winocour, P.H., Ishola, M., Durrington, P.N. & Anderson,D.C. Lipoprotein abnormalities in insulin-dependent dia-betes mellitus. Lancet 1986, i: 1176- 1178.

57. Vannini, P., Ciavarella, A., Flammini, M. et al. Lipidabnormalities in insulin-dependent diabetic patients withalbuminuria. Diabetes Care 1984, 7: 151-154.

58. Jones, S.L., Close, C.F., Mattock, M.B., Jarrett, R.J., Keen,H. & Viberti, G.C. Plasma lipid and coagulation factorconcentrations in insulin dependent diabetics with micro-albuminuria. Br Med J 1989, 298: 487-490.

59. Borch-Johnsen, K., Anderson, P.K. & Deckert, T. The effectof proteinuria on relative mortality in type 1 (insulin-dependent) diabetes mellitus. Diabetologia 1985, 28: 590-596.

60. Borch-Johnsen, K. & Kreiner, S. Proteinuria: value as apredictor of cardiovascular mortality in insulin dependentdiabetes mellitus. Br Med J 1987, 294: 1651-1654.

61. Winocour, P.H., Durrington, P.N., Ishola, M., Anderson,D.C. & Cohen, H. Influence of proteinuria on vasculardisease, blood pressure, and lipoproteins in insulin-depen-dent diabetes mellitus. Br Med J 1987, 294: 1648-1651.

62. Bierman, E.L., Amaral, A.P. & Belknap, B.H. Hyper-lipidaemia and diabetes mellitus. Diabetes 1966, 15: 675-679.

63. Bagdade, J.D., Porte, D. Jr & Bierman, E.L. Diabeticlipaemia: a form of acquired fat-induced lipaemia. N Engl JMed 1967, 276: 427-433.

64. Weidman, S.W., Ragland, J.B., Fisher, J.N., Kitabchi, A.E.& Sabesin, S.M. Effects of insulin on plasma lipoproteins indiabetic ketoacidosis: evidence for a change in high densitylipoprotein composition during treatment. J Lipid Res 1982,23: 171-182.

65. Lopes-Virella, M.F., Wohltmann, H.J., Mayfield, R.K.,Loadholt, C.B. & Colwell, J.A. Effect ofmetabolic control onlipid, lipoprotein, and apolipoprotein levels in 55 insulin-dependent diabetic patients. A longitudinal study. Diabetes1983, 32: 20-25.

66. Pietri, A., Dunn, F.L. & Raskin, P. The effect of improveddiabetic control on plasma lipid and lipoprotein levels: acomparison of conventional therapy and continuous sub-cutaneous insulin infusion. Diabetes 1980, 29: 1001-1005.

67. Falko, J.M., O'Dorisio, T.M. & Cataland, S. Improvement ofhigh density lipoprotein-cholesterol levels. Ambulatory typeI diabetics treated with the subcutaneous insulin pump.JAMA 1982, 247: 37-39.

68. Dunn, F.L., Pietri, A. & Raskin, P. Plasma lipid andlipoprotein levels with continuous subcutaneous insulininfusion in type I diabetes mellitus. Ann Intern Med 1981, 95:426-431.

69. James, R.W. & Pometta, D. Differences in lipoproteinsubfraction composition and distribution between type 1diabetic men and control subjects. Diabetes 1990, 39:1158-1164.

70. Kaufmann, R.L., Assal, J.Ph., Soeldner, J.S. et al. Plasmalipids in diabetic children. Effect of diet restricted incholesterol and saturated fats. Diabetes 1975, 24: 672-679.

71. Hollenbeck, C.B., Connor, W.E., Riddle, M.C., Alaupovic,P. & Lecklem, J.E. The effects of a high-carbohydrate,low-fat, cholesterol restricted diet on plasma lipids, lipopro-tein and apoprotein concentrations in insulin-dependent(type 1) diabetes mellitus. Metabolism 1985, 34: 559-566.

72. Stone, D. & Connor, W.E. The prolonged effects of alow-cholesterol, high carbohydrate diet upon serum lipids indiabetic patients. Diabetes 1963, 12: 127-132.

73. Jenkins, D.J.A., Wolever, T.M.S., Taylor, R.H., Reynolds,D., Nineham, R. & Hockaday, T.D.R. Diabetic glucosecontrol, lipids and trace elements on long term guar. Br MedJ1980, 280: 1353-1354.

74. Sosenko, J.M., Breslow, J.L., Miettinen, O.S. & Gabbay,K.H. Hyperglycaemia and plasma lipid levels. A prospectivestudy of young insulin-dependent diabetic patients. N Engi JMed 1980, 302: 650-654.

75. Ostlund, R.E., Semenkovich, C.F. & Schechtman, K.B.Quantitative relationship between plasma lipids and glyco-hemoglobin in type I patients. Longitudinal study of 212patients. Diabetes Care 1989, 12: 332-336.

76. Elkeles, R.S., Wu, J. & Hambley, J. Haemoglobin Al, bloodglucose and high density lipoprotein cholesterol in insulin-requiring diabetics. Lancet 1978, i{: 547-549.

77. Gonen, B., White, N., Schonfeld, G., Skor, D., Miller, P. &Santiago, J. Plasma levels of apoprotein B in patients withdiabetes mellitus: the effect of glycaemic control. Metabolism1985, 34: 675-679.

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from

DIABETES: TREATMENT AND SERUM LIPIDS 937

78. Lopes-Virella, M.F., Wohltmann, H.J., Loadholt, C.B. &Buse, M.G. Plasma lipids and lipoproteins in young insulin-dependent diabetic patients: relationship with control. Dia-betologia 1981, 21: 216-223.

79. Semenkovitch, C.F., Ostlund, R.E. & Schechtman, K.B.Plasma lipids in patients with type I diabetes mellitus.Influence of race, gender, and plasma glucose controls: Lipidsdo not correlate with glucose control in black women. ArchIntern Med 1989, 149: 51-56.

80. Winocour, P.H. & Laker, M.F. Drug therapy for diabeticdyslipoproteinaemia: a practical approach. Diabetic Med1990, 7: 292-298.

81. Winocour, P.H., Durrington, P.N., Bhatnagar, D. et al.Double blind placebo-controlled study on the effects ofbezafibrate on blood lipids, lipoproteins, and fibrinogen inhyperlipidaemic type I diabetes mellitus. Diabetic Med 1990,7: 736-743.

82. Editorial. Fish oils and diabetic microvascular disease.Lancet 1990, 1: 508-509.

83. Jensen, T., Stender, S., Goldstein, K. et al. Partial normalisa-tion by dietary cod liver oil of decreased microvascularalbumin leakage in patients with insulin-dependent diabetesand albuminuria. N Engl J Med 1989, 321: 1572-1577.

84. Mori, T.A., Vandongen, R. & Maserei, J.R.L. Fish oil-induced changes in apolipoproteins in IDDM subjects.Diabetes Care 1990, 13: 725-732.

85. Bagdade, J.D., Buchanan, W.E., Levy, R.A., Subbaiah, P.V.& Ritter, M.C. Effects of w-3 fish oils on plasma lipids,lipoprotein composition and postheparin lipoprotein lipasein women with IDDM. Diabetes 1990, 39: 426-431.

86. Stacpoole, P.W., Alig, J., Ammon, L. & Crockett, S.E.Dose-response effects of dietary marine oil on carbohydrateand lipid metabolism in normal subjects and patients withhypertriglyceridaemia. Metabolism 1989, 38: 946-956.

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from

lipoproteins.effect on serum lipids and Treatment of diabetes: the

P. K. Merrin and R. S. Elkeles

doi: 10.1136/pgmj.67.792.9311991 67: 931-937 Postgrad Med J

tationhttp://pmj.bmj.com/content/67/792/931.cifound at: Updated information and services can be

These include:

serviceEmail alerting

top right corner of the online article. cite this article. Sign up in the box at the Receive free email alerts when new articles

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:

group.bmj.com on February 13, 2018 - Published by http://pmj.bmj.com/Downloaded from