What should we measure in the diabetic patient and how does this respond to therapy? (2024)

Introduction

Diabetes is characterized by hyperglycaemia and by the development of complications. The criteria for diagnosis have been revised and published by the World Health Organization (WHO). In a symptomatic patient, a random plasma glucose concentration of ≥11.1 mmol l⁻¹ is diagnostic. For asymptomatic patients, two diagnostic tests are required. A fasting glucose level ≥7 mmol l⁻¹ confirms the diagnosis (providing this is supported by a second diagnostic test). The 75 g oral glucose tolerance test, as recommended by the WHO, is used for diagnosis in borderline cases or for academic studies. A plasma glucose concentration ≥11.1 mmol l⁻¹ at 2 h following oral glucose confirms the diagnosis (Table 1).

Table 1.

Values for diagnosis of diabetes mellitus and other categories of hyperglycaemia.

	Glucose concentration (mmol l⁻¹)
	Whole blood Venous	Capillary	Plasma Venous
Diabetes mellitus
Fasting or	≥6.1	≥6.1	≥7.0
2 h post glucose load	≥10.0	≥11.1	≥11.1
Impaired glucose tolerance (IGT)
Fasting (if measured) and	≥6.1 and	<6.1 and	<7.0 and
2 h post glucose load	≥6.7	≥7.8	≥7.8
Impaired fasting glycaemia (IFG)
Fasting	≥5.6 and	≥5.6 and	≥6.1 and
	<6.1	<6.1	<7.0
and (if measured)
2 h post glucose load	<6.7	<7.8	<7.8

Open in a new tab

The microvascular complications specific to diabetes include retinopathy, nephropathy and neuropathy. Patients with diabetes are predisposed also to macrovascular disease (coronary heart disease, stroke and peripheral vascular disease). The macrovascular complications are not specific but occur with increased frequency in the diabetic compared with the nondiabetic population [1, 2]. Mortality from coronary heart disease (CHD) is 2–4 times higher than in the nondiabetic community. CHD is a problem for all diabetic patients but particularly for South Asian patients with diabetes whereas stroke is a problem especially for the African-Caribbean diabetic population [3–9].

Diabetes is a common disorder. Type 2 diabetes is much more common than Type 1 (insulin-dependent) and our discussion is confirmed to Type 2. In the adult population, the prevalence is 3–5% and in the population >65 years, this rises to 10%. It is commoner in non-European people in the UK than in Europeans (this ethnic difference is evident elsewhere). In the South Asian population >40 years, for example, the prevalence is 20–30%. Diabetes is also becoming more common and by the year 2010, more than 200 million people worldwide will have the disorder. Currently, diabetes consumes 5–15% of healthcare budgets in the developed world and most of this cost is due to the treatment of diabetic complications. As diabetes becomes more common in the developing world [10], it will have major implications for public health services and their costs.

Lesser degrees of abnormal glucose tolerance

Impaired fasting glycaemia (IFG) refers to those in the population with a fasting plasma venous glucose concentration ≥6.1<7.0 mmol l⁻¹. If an oral glucose tolerance test is performed, subjects with a 2 h plasma venous glucose level ≥7.8<11.1 mmol l⁻¹ are classified as having impaired glucose tolerance (IGT). The importance of IFG and IGT is that these subjects, while not predisposed to the microvascular complications of diabetes, are at risk of macrovascular disease [11–13]. People with lesser degrees of glucose elevation frequently demonstrate other CHD risk factors, such as insulin resistance, central obesity, hypertension, elevated triglyceride concentrations and a reduction in HDL cholesterol [9, 14–16]. The Metabolic Syndrome (Syndrome X, Reaven's Syndrome) has been defined as abnormal glucose tolerance in association with insulin resistance, together with two or more of the following: blood pressure ≥160/90 mm Hg; triglyceride ≥1.7 and/or HDL cholesterol ≤0.9 mmol l⁻¹; waist: hip ratio ≥0.90 and/or BMI ≥30 kg m⁻²; microalbuminuria ≥20 µg min⁻¹. Patients with lesser degrees of glucose elevation, with or without the other components of the Metabolic Syndrome, are at risk of developing frank diabetes. The frequency of progression from IGT to diabetes, for example, ranges from 2% to 14% per year, depending on the ethnic origin of the population. IGT is a risk factor for CHD however, even independently of the development of diabetes, and many patients will die of CHD before becoming diabetic.

Measurement in diabetes

Some of the measurements which are performed routinely in all patients with diabetes, and others which are reserved for research, are listed in Table 2.

Table 2.

Measurements in diabetes.

1. Plasma and urine glucose

2. Glycosolated haemoglobin

3. Lipids (total and HDL choolesterol, triglycerides)

4. Creatinine

5. Urinary albumin

6. Clinical measurements (anthropometry, blood pressure)

7. Electrocardiogram

8. Others

non-invasive vascular assessments

markers of vascular function

markers of vascular re-modelling

thrombotic thrombolytic markers

Open in a new tab

Urine glucose measurements

Measurement of glucose in urine has been possible for patients at home since the 1940s [reviewed in 17]. The initial Clinitest reagent tablets (a modification of Benedict's test) required the addition of the tablet in a tube. The chemical reaction caused urine glucose to be oxidized, reducing cupric sulphate to a colour ranging from blue to orange. This permitted semiquantitative results to be obtained by visual comparison with a colour chart.

Reagent tablets were replaced by test strips in the 1950s. The Clinistix reagent strips were the first to be introduced and utilized glucose oxidase and peroxidase with a chromogen, covered by a semi-permeable membrane. A colour change was produced as glucose was oxidized. Again, a semi-quantitative result was obtained by inspection with a colour chart.

Urine glucose testing systems remain widely used throughout the world. They have the advantage of being inexpensive and some patients find urine testing more acceptable than blood testing. They do rely on the renal threshold for glucose being constant (and there is some variability in this). Urine testing cannot help with the detection of hypoglycaemia.

Measurements of circulating glucose

Home blood glucose monitoring became possible in the 1960s and this is the method most frequently utilized for home monitoring of diabetic control in the Western world. It is employed by most patients treated with insulin and by some Type 2 diabetic patients managed on diet with or without oral hypoglycaemic therapy. Blood glucose profiles, with glucose measurements extending from before breakfast until prebed, are useful to guide patients as to their optimum insulin therapeutic regimen. Typically, measurements are made on 2–6 occasions through the sampling period. Knowledge of the kinetics of insulin action, together with these glucose measurements, provides the basis for therapeutic adjustments. During intercurrent illness, patients are encouraged to test more frequently. It is usual for insulin requirements to increase at such times and knowledge of ambient glucose levels may permit satisfactory diabetic control to be obtained throughout the illness. Blood glucose measurements also provide information about hypoglycaemia. Such information is extremely useful to patients and their physicians when there is any doubt as to the cause of possible hypoglycaemic symptoms.

Early test strips, such a Dextrostix, were semi-quantitative. They were read visually and were very useful. Their reliability did require satisfactory vision (which may not be the case in some diabetic patients) and required that the patient was capable of timing accurately the measurement after application of the blood to the test strip. The methods also required some manual dexterity, for example to satisfactorily remove red cells from the strip surface [18].

For this reason, other glucose measurement systems have been introduced. Reflectance meters were introduced at the end of the 1960s and through the 1970s. These meters measure light reflected from a coloured test strip, providing quantitative glucose estimations. Early machines were cumbersome and needed skilled usage. They required manual calibration and were sensitive to incorrect application of blood to the test strip. Recent systems are portable and compact, and require much less operator skill. Measuring the colour on the underside of the strip has removed the requirement for correct wiping technique at the end of the reaction period. Timing is automatic and there are in-built safeguards to warn that meter maintenance is required.

Absorbance meters measured the absorption of light at a specific wavelength. The recently introduced HemoCue B system is a glucose home monitoring system which utilizes absorbance methodologies. It is one of the most accurate of all home glucose measurement devices.

Biosensors became available in the late 1980s and into the 1990s. One of these, produced by MediSense, utilized glucose oxidase with ferrocene as the mediator. When glucose was oxidized to gluconolactone, the electrons released were absorbed by ferrocene to form ferrocinium, which was oxidized at an electrode. The current produced was proportional to the glucose concentration. These sensors were robust, as the electrodes to which blood was applied were external to the body of the equipment, and again effective blood wiping was not required.

Glycosylated haemoglobin analysis

Circulating haemoglobin in normal adults comprises HbA, HbA₂ and HbF (Figure 1). These three haemoglobin variants share an α chain, with specificity dictated by the non-α chain.

Figure 1.

1. Chromatography

1a. Cation exchange chromatography

Glycation of haemoglobin results in a small change in electrical charge of the haemoglobin molecule. This alters its migration in electric fields or cation exchange resins. Early column cation exchange methods suffered from poor reproducibility, interference by aldimine in blood and by lipids in lipaemic serum. Later methods, e.g. the Bio-Rad Laboratories HbA_1c Bicolumn Test, managed to overcome many of these problems and were widely adopted.

1b. High performance liquid chromatography (h.p.l.c)

Excellent cation exchange resin h.p.l.c. methods appeared later and provided rapid and reproducible assays which could be automated. In recent years, these methods have been automated, permitting high throughput as demand for HbA_1c measurement has grown. Low pressure and high pressure systems have been devised with coefficients of variation (C.V.) of <3%.

1c. Electrophoresis

The very small change in isoelectric point following glycosylation of haemoglobin (0.01 pH units) has required the development of specialized techniques. Agar gel electroendosmosis with detection of the separated bands by scanning densitometry provide reasonable quantification. This principle was adopted by a commercial kit (CIBA Corning) which was widely employed in the 1980s. This method has been largely replaced by more automated techniques. In recent years, improved electrophoretic methods capable of automation have been developed and are now widely employed (e.g. Diatrac; Beckman).

1d. Affinity chromatography and affinity separation

The principle of this method is the binding of 1,2-cis diol groups on glycosylated haemoglobin to immobilized boronate, often on agarose gel. Mutual affinity chromatography methods have performed well in comparison with other methods. Inter-laboratory agreement has been less satisfactory, which might be an issue for external audit. Some variation in ligand concentration between batches may cause significant variations in results.

Affinity h.p.l.c. methods have been developed to increase throughput and measure glycosylated haemoglobin with a C.V. of <3%. Conversion from total glycosylated haemoglobin to HbA_1c values is necessary using a linear regression correlation. Automated methods (e.g. Abbott IMx) have been developed to give reproducible results and rapid analyses using disposable cartridges, permitting measurements in clinic (e.g. Glycosal Analyser, Provalis).

1e. Immunoassay methodologies

An early enzyme immunoassay utilized a monoclonal antibody which detected the first 8 amino acids on the β-chain of haemoglobin, together with the attached glucose. The method is reproducible and correlates well with other methods but the reference range was lower than for other HbA_1c methodologies.

An immunoturbidimetric method utilizing an antibody which recognized the first four amino acids and glucose of the β-chain has proven robust (Roche Diagnostics Systems). A similar method employing proteolytic degradation to render the β-N terminal structure more accessible has been devised. A latex agglutination method, with the particles coated with HbA_1c-specific monoclonal antibodies, has given good precision and correlation with other methodologies. This method has also been capable of in-clinic use.

1f. General issues

Whichever method of glycosylated haemoglobin measurement is employed, its specificity is important. Measurement of HbA_1c, aligned to the Diabetes Control and Complication Trial (DCCT) assay or to its successor from the International Federation of Clinical Chemistry, is essential [20]. HbA_1c estimations will be utilized in future in auditing diabetic control between centres and standardization of results is essential for this purpose despite the fact that other factors influence the concentration [21].

The importance of glycosylated haemoglobin measurement results from the fact that they are uninfluenced by recent food ingestion and that they reflect integrated glucose concentrations over time. HbA_1c reflects complication risk in the general population and in diabetic patients. In population studies, the risk of retinopathy increases with HbA_1c concentrations between 6 and 7%, and is very low in those with values below 6%. Similarly, the risk of heart disease increases with increasing HbA_1c levels, even in the general population, as illustrated in a recent study in Norfolk, UK [22]. In this Norfolk Study, the risk of death from CHD was increased 5-fold in those with an HbA_1c >6.9% when compared with those in whom the HbA_1c level was <5%. In patients with established Type 2 diabetes, the incidence of myocardial infarction increases with increasing HbA_1c level, as does the incidence of microvascular end points [23].

1g. Blood pressure

Hypertension occurs with greater frequency in Type 2 diabetes than in the normal population [24]. Blood pressure, like blood glucose control, is predictive of cardiovascular events [25]. Effective treatment of blood pressure reduces the risk [26, 27]. In the diabetic subgroups of large blood pressure-lowering trials, the benefits of hypotensive treatment have been at least as good in diabetic patients as in the nondiabetic population. Cardiovascular events decreased in the diabetic patients, for example, by 51% in the HOT Study [28]. Reductions in the incidence of stroke and possibly myocardial infarction were noted in the United Kingdom Prospective Diabetes Study [29, 30].

1h. Renal function and urinary albumin

Diabetic nephropathy is characterized by persistent proteinuria, declining glomerular function and hypertension [31]. It is also associated with an increase in CHD risk. Nephropathy develops in both Type 1 and Type 2 diabetes. In Type 1 diabetes, the prevalence is 15–20% with a peak incidence after 15–16 years of diabetes. In Type 2 diabetes, prevalence rates of proteinuria have varied from 3% to 16%. The natural history of nephropathy in Type 2 diabetes is less well understood but, in view of the high prevalence of Type 2 diabetes compared with Type 1, almost 50% of all diabetic patients requiring renal replacement therapy have Type 2 disease.

Clinical proteinuria is defined as urinary protein excretion of >500 mg day⁻¹, equivalent to an albumin excretion of 300 mg day⁻¹ (200 µg min⁻¹). This phase of clinical proteinuria is preceded by a phase of microalbuminuria, when protein excretion is below the threshold of detection by Albustix, and must be measured by more sensitive methods. In Type 1 diabetes, 15–28% of patients have persistent microalbuminuria and in Type 2 patients, the proportion has been 15–59% in different series.

The importance of proteinuria, as stated above, relates not merely to renal function but also to CHD and the relative cardiovascular mortality is increased 10-fold in patients with clinical nephropathy. CHD risk is increased even at the phase of microalbuminuria.

1i. Dyslipidaemia

The typical dyslipidaemia of Type 2 diabetes is an increase in circulating triglyceride and a decrease in HDL cholesterol levels. The severity of the dyslipidaemia is predictive of vascular disease. Although total and LDL cholesterol concentrations may be in the range for the nondiabetic community, their concentrations correlate with CHD in Type 2 diabetic patients [32]. The relative severity of the triglyceride elevation and the HDL cholesterol suppression may be greater in women than in men [33].

Outcome data have shown the benefit of treatment of the dyslipidaemia. Early studies showed a reduced CHD event rate after 3 years of therapy with bezafibrate of dyslipidaemic patients with Type 2 diabetes (34). In the 4S Study, Simvastatin produced a 55% decrease and in the CARE Study, Pravastatin produced a 25% decrease in CHD events [35, 36].

Conclusions

We have focused on a small number of measurements which are widely performed. It is evident (Table 2) that others may become relevant. Certain patients develop CHD or microvascular disease quickly whereas others with apparently similar blood glucose and blood pressure control do not suffer complications over many years. The challenge is to find ways to distinguish those who require intensive risk factor intervention from those who are at lower risk. This is likely to involve biochemical and genetic approaches, and will ensure an exciting period of intensive research over the next decade.

References

1.Pyorala K, Laakso M, Uusitupa M. Diabetes and atherosclerosis: an epidemiologic view. Diabetes Metab Rev. 1987;3:463–524. doi: 10.1002/dmr.5610030206. [DOI] [PubMed] [Google Scholar]
2.Panzram G. Mortality and survival in type 2 (non-insulin-dependent) diabetes mellitus. Diabetologia. 1987;30:123–131. doi: 10.1007/BF00274216. [DOI] [PubMed] [Google Scholar]
3.Simmons D, Williams DRR, Powell MJ. The Coventry Diabetes Study. Prevalence of diabetes and impaired glucose tolerance in Europids and Asians. Quart J Med. 1991;81:1021–1030. doi: 10.1093/qjmed/81.3.1021. [DOI] [PubMed] [Google Scholar]
4.Dowse GK, Zimmet PZ. The prevalence and incidence of non-insulin-dependent diabetes mellitus. In: Alberti KGMM, Mazze RS, editors. Current Research in Non-Insulin-Dependent Diabetes. Amsterdam: Elsevier; 1989. pp. 37–60. [Google Scholar]
5.King H, Rewers M. Global estimates for prevalence of diabetes mellitus and impaired glucose tolerance in adults. WHO Ad Hoc Diabetes Reporting Group Diabetes Care. 1993;16:157–177. doi: 10.2337/diacare.16.1.157. [DOI] [PubMed] [Google Scholar]
6.Chaturvedi N, McKeigue PM, Marmot MG. Relationship of glucose intolerance to coronary risk in Afro-Caribbeans compared with Europeans. Diabetologia. 1994;37:765–772. doi: 10.1007/BF00404333. [DOI] [PubMed] [Google Scholar]
7.Unwin N, Alberti KGMM, Bhopal R, Harland J, Watson W, White M. Comparison of the current WHO and new ADA criteria for the diagnosis of diabetes mellitus in three ethnic groups in the UK. Diabet Med. 1998;15:554–557. doi: 10.1002/(SICI)1096-9136(199807)15:7<554::AID-DIA626>3.0.CO;2-E. [DOI] [PubMed] [Google Scholar]
8.Harris MI, Hadden WC, Knowler WC, Bennett PH. Prevalence of diabetes and impaired glucose tolerance and plasma glucose levels in U.S. population aged 20–74 yr. Diabetes. 1987;36:523–534. doi: 10.2337/diab.36.4.523. [DOI] [PubMed] [Google Scholar]
9.McKeigue PM, Shah B, Marmot MG. Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet. 1991;337:382–386. doi: 10.1016/0140-6736(91)91164-p. [DOI] [PubMed] [Google Scholar]
10.ADA. Alexandra, VA: American Diabetes Association; 1993. Direct and indirect costs of diabetes in the USA in 1992. [Google Scholar]
11.Alberti KGMM. The clinical implications of impaired glucose tolerance. Diabet Med. 1996;13:927–937. doi: 10.1002/(SICI)1096-9136(199611)13:11<927::AID-DIA279>3.0.CO;2-E. [DOI] [PubMed] [Google Scholar]
12.Eschwege E, Richard JL, Thibult N, et al. Coronary heart disease mortality in relation with diabetes, blood glucose and plasma insulin levels. The Paris Prospective Study, ten years later. Horm Metab Res Suppl. 1985;15:41–46. [PubMed] [Google Scholar]
13.Gabir MM, Hanson RL, Dabelea D, et al. Plasma glucose and prediction of microvascular disease and mortality: evaluation of 1997 American Diabetes Association and 1999 World Health Organization criteria for diagnosis of diabetes. Diabetes Care. 2000;23:1113–1118. doi: 10.2337/diacare.23.8.1113. [DOI] [PubMed] [Google Scholar]
14.Ohlson LO, Larsson B, Svardsudd K, et al. The influence of body fat distribution on the incidence of diabetes mellitus. 13.5 years of follow-up of the participants in the study of men born in 1913. Diabetes. 1985;34:1055–1058. doi: 10.2337/diab.34.10.1055. [DOI] [PubMed] [Google Scholar]
15.Mather HM, Chaturvedi N, Kehely AM. Comparison of prevalence and risk factors for microalbuminuria in South Asians and Europeans with type 2 diabetes mellitus. Diabet Med. 1998;15:672–677. doi: 10.1002/(SICI)1096-9136(199808)15:8<672::AID-DIA648>3.0.CO;2-3. [DOI] [PubMed] [Google Scholar]
16.Drury DL. Hypertension. Clin Endocrinol Metab. 1988;2:375–379. doi: 10.1016/s0950-351x(88)80038-7. [DOI] [PubMed] [Google Scholar]
17.Metcalfe K. Monitoring the patient with diabetes in clinic. In: John WG, editor. Monitoring Glycaemic Control in the Diabetic Patient. Harcourt Health Communications; 2000. pp. 55–64. [Google Scholar]
18.Worth RC, Harrison K, Anderson J, Johnston DG, Alberti KGMM. A comparison of blood glucose test strips. Diabetes Care. 1981;4:407–411. doi: 10.2337/diacare.4.3.407. [DOI] [PubMed] [Google Scholar]
19.John WG. Glycated haemoglobin analysis. In: John WG, editor. Monitoring Glycaemic Control in the Diabetic Patient. Harcourt Health Communications; 2000. pp. 79–102. [Google Scholar]
20.Marshall SM, Barth JH. Standardization of HbA1c measurements – a consensus statement. Diabetic Med. 2000;17:5–6. doi: 10.1046/j.1464-5491.2000.00228.x. [DOI] [PubMed] [Google Scholar]
21.Kilpatrick ES, Maylor PW, Keevil BG. Biological variation of glycated hemoglobin. Implications for diabetes screening and monitoring. Diabetes Care. 1998;21:261–264. doi: 10.2337/diacare.21.2.261. [DOI] [PubMed] [Google Scholar]
22.Khaw KT, Wareham N, Luben R, et al. Glycated haemoglobin, diabetes, and mortality in men in Norfolk cohort of European Prospective Investigation of Cancer and Nutrition (EPIC-Norfolk) Br Med J. 2001;322:15–18. doi: 10.1136/bmj.322.7277.15. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Stratton IM, Adler AI, Neil HA, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. Br Med J. 2000;321:405–412. doi: 10.1136/bmj.321.7258.405. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.UKPDS. The Hypertension in Diabetes Study Group. Hypertension in diabetes study (HDS). I. Prevalence of hypertension in newly presenting type 2 diabetic patients and the association with risk factors for cardiovascular and diabetic complications. J Hypertens. 1993;11:309–317. doi: 10.1097/00004872-199303000-00012. [DOI] [PubMed] [Google Scholar]
25.UKPDS. The Hypertension in Diabetes Study Group. Hypertension in diabetes study (HDS). II. Increased risk of cardiovascular complications in hypertensive type 2 diabetic patients. J Hypertens. 1993;11:319–325. doi: 10.1097/00004872-199303000-00013. [DOI] [PubMed] [Google Scholar]
26.Curb JD, Pressel SL, Cutler JA, et al. Effect of diuretic-based antihypertensive treatment on cardiovascular disease risk in older diabetic patients with isolated systolic hypertension. JAMA. 1996;276:1886–1892. [PubMed] [Google Scholar]
27.HOPE study. Effects of ramipril on cardiovascular and microvascular outcomes in people with diabetes mellitus: results of the HOPE study and MICRO-HOPE substudy. Heart Outcomes Prevention Evaluation Study Investigation. Lancet. 2000;355:253–259. [PubMed] [Google Scholar]
28.Hansson L, Zanchetti A, Carruthers SG, et al. Effects of intensive blood pressure lowering and low-dose aspirin in patients with hypertension: principal results of the hypertension optimal treatment (HOT) randomised trial. Lancet. 1998;351:1755–1762. doi: 10.1016/s0140-6736(98)04311-6. [DOI] [PubMed] [Google Scholar]
29.UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33) Lancet. 1998;352:837–853. [PubMed] [Google Scholar]
30.UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes. UKPDS 38. Br Med J. 1998;317:703–713. [PMC free article] [PubMed] [Google Scholar]
31.Estacio RO, Schrier RW. Diabetic nephropathy. Pathogenesis, diagnosis, and prevention of progression. Adv Intern Med. 2001;46:359–408. [PubMed] [Google Scholar]
32.Howard BV, Robbins DC, Sievers ML, et al. LDL cholesterol as a strong predictor of coronary heart disease in diabetic individuals with insulin resistance and low LDL. The Strong Heart Study. Arterioscler Thromb Vasc Biol. 2000;20:830–835. doi: 10.1161/01.atv.20.3.830. [DOI] [PubMed] [Google Scholar]
33.Walden CE, Knopp RH, Wahl PW, Beach KW, Strandness E., Jr Sex differences in the effect of diabetes mellitus on lipoprotein triglyceride and cholesterol concentrations. N Engl J Med. 1984 October 11;311(15):953–959. doi: 10.1056/NEJM198410113111505. [DOI] [PubMed] [Google Scholar]
34.Elkeles RS, Diamond JR, Poulter C, et al. Cardiovascular outcomes in type 2 diabetes. A double-blind placebo-controlled study of bezafibrate: the St Mary's, Ealing, Northwick Park Diabetes Cardiovascular Disease Prevention (SENDCAP) Study. Diabetes Care. 1998;21:641–648. doi: 10.2337/diacare.21.4.641. [DOI] [PubMed] [Google Scholar]
35.Sacks FM, Pfeffer MA, Moye LA, et al. The effects of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med. 1996;335:1001–1009. doi: 10.1056/NEJM199610033351401. [DOI] [PubMed] [Google Scholar]
36.Sacks FM, Tonkin AM, Shepherd J, et al. Effect of pravastatin on coronary disease events in subgroups defined by coronary risk factors: the Prospective Pravastatin Pooling Project. Circulation. 2000;102:1893–1900. doi: 10.1161/01.cir.102.16.1893. [DOI] [PubMed] [Google Scholar]