This lecture focused on stroke, which appears to be increased 2-fold in diabetic men and 4-fold in diabetic women according to data from the Framingham Heart Study. Overall, women suffer more CV complications from diabetes than men; the CV protective effect typically seen in women over men is lost in diabetic women.

Several studies have now shown that when the other risk factors (hypercholesterolemia, hypertension, obesity, smoking) are taken into account, the relative risk for cardiovascular disease (CVD) for diabetes change little. This includes all CVD, coronary heart disease (CHD), and stroke. The relative risk (RR) for stroke is larger in diabetics (3.0-3.7 in men, 2.8-3.7 in women) than is the relative risk for CHD (2.0-2.4 in men, 3.2-3.6 in women) in this group. The newly diagnosed diabetic has the same poor prognosis as those persons with diabetes for a number of years_meaning that asymptomatic, undetected diabetes in the community is a very important public health problem.

Relative risk for stroke

Diabetes is one of the strongest risk factors for stroke, as shown in Finnish study in several thousand people. From this study data it can be speculated that some of the stroke risk attributed to hypertension could actually be due to undetected diabetes in persons with hypertension. In men, diabetes confers a 3.35% risk and in women 4.89%. In comparison, smoking conferred a 1.48 risk in men and 2.04 risk in women; cholesterol a 1.19 risk in men 1.19 and 0.94 risk in women; systolic blood pressure 1.02 in men and 1.01 risk in women.

The effect of age is about the same when comparing diabetic and non-diabetic men who developed stroke. Smoking is a risk factor for stroke only in the non-diabetic men, with a relative risk of 1.80 compared to 0.81 in diabetics. The effect of serum cholesterol, systolic blood pressure (SBP), body mass index and antihypertensive drug treatment is about the same. In women, smoking is a risk factor for stroke only in the non-diabetics compared to diabetics. This is likely due to diabetics ceasing smoking upon the advice of their physicians when diagnosed with diabetes. The effect of SBP is the same in diabetic and non-diabetic women, with a relative risk of 1.01.

Fatal and non-fatal stroke events

The 7-year incidence of fatal and non-fatal stroke events is greater in diabetics with high glucose levels (> 13 mmol/L) compared to diabetics with low glucose levels (< 13 mmol/L) in the Finnish study. There is some modification of the risk with high triglycerides versus low triglycerides, low HDL versus high HDL, and high total cholesterol versus low total cholesterol. But, the main effect is due to the high glucose levels.

For cerebral infarction there is a 2- to 5-fold increase in the risk of stroke in diabetes, whereas in hemorrhagic stroke there appears to be no increase with diabetes. One study has indicated that subarachnoid hemorrhage may be decreased in patients with Type I diabetes. A study from the US shows that hemorrhagic stroke is not increased with the glucose level in the population. However, these glucose levels are not very high, with the highest level about 189 mg/dl. Thromboembolic stroke is increased with increased levels of blood glucose.

Diabetes is a risk factor for ischemic stroke and intracerebral hemorrhage (ICH). This large study (nearly 800 ICH strokes and 2400 ischemic strokes) is the first demonstration of this effect of diabetes and is in contrast to previous beliefs. Smoking is not related to ICH.

A prospective study from Chicago showed there was little age-specific relation between the postload plasma glucose and risk of various CVD, CHD, stroke and all-cause mortality. For CHD the effect was stronger in younger men (age 40-59) versus older men (60-74), and was the opposite in women. For stroke the effect of high plasma glucose was the same in both groups in men (1.19 RR older, 1.23 RR younger). For women, however, the risk of stroke was increased in the older age group (1.17 RR, vs 1.01 RR).

Different pathologies in diabetics compared to non-diabetics may exist. A recent Japanese study of autopsies with verified cerebrovascular disease interestingly showed there were non-fatal clinical manifestations more commonly in diabetics than in non-diabetics (68% vs 23% in men, 50% vs 27% in women). The frequency of small infarctions was also greater in diabetics (76% vs 40% in men, 73% vs 57% in women). The number of multiple small infarcts in the brain was also higher in the diabetics (65% vs 23% in men, 625 vs 47%).

The Japanese study also looked at the magnitude of stroke by measuring the lesion volumes by CT scan in relation to plasma glucose, serum cortisol, plasma insulin, plasma noradrenaline, and hemoglobin A1C. Plasma glucose was strongly associated with the presence of a brain lesion, with a correlation coefficient of 0.469. Serum cortisol had a correlation coefficient of 0.542 and plasma insulin 0.399, both part of the metabolic syndrome.

A Finnish 6-year follow-up study after myocardial infarction (MI) showed that a history of a stroke conferred a 3.49 RR for a subsequent stroke, while diabetes had a RR of 2.20. Diabetes was a stronger risk factor for stroke than atrial fibrillation with a 1.63 RR.

Fasting blood glucose (FBG) was used to estimate stroke risk in a population-based study of middle-aged people in Scotland. Very low glucose levels had an increased risk of stroke, but thereafter there was a nearly linear association. The age-adjusted stroke rate was 30.9 for FBG < 4.2 in the 20-yr follow-up, 21.4 for a FBG of 4.3-4.6 with linear increases thereafter. Although this was a trend, it did not reach statistical significance.

The DECODE study was a pooled analysis of 10 studies from seven European countries with 15,388 men and 726 women. During the median 8.8 years follow-up, there were 4061 all-cause deaths and 1366 CVD deaths.

For patients with IGT (7.8-11.0 mM, 2-hr GTT) there was a 26% increase in stroke mortality, 34% increased risk of CVD mortality, a 28% increase in CHD mortality, and a 40% increase in all-cause mortality. Comparing diabetics (> 11.1 mM, 2 hr GTT) to non-diabetics, there was a 74% increase for stroke mortality, a 55% increase for CVD mortality, a 64% increase in CHD mortality, and a 92% increase in all-cause mortality.

More extensive data analysis shows an increase in stroke from very low levels of 2-hr plasma glucose. A linear trend for mortality related to glucose levels was also seen: 3.07 hazard ratio (HR) for a plasma glucose of 3.5, 5.82 HR for plasma glucose of 10.5, 6.48 HR for plasma glucose of 12.5, and 5.85 HR for a plasma glucose > 14.0.

The way IGT is measured may be important because a comparison of FBG levels showed an increased risk of stroke that was not as visible with the 2-hour plasma glucose test. Glucose tolerance tests may need to be used to identify those persons at risk of CVD due to moderately high glucose levels.

Recent studies show a strong positive relation between the presence of Type I diabetes and the risk of stroke, in contrast to earlier studies. More than 50% of type I diabetes is adult-onset. In Type II diabetes, the association to stroke is strong. Type II often starts in the elderly and has a very long asymptomatic period during which CV problems develop.

Diabetic nephropathy plays a very important role in CV morbidity. Patients with high levels of proteinuria, particularly women, have increased risk of heart failure, stroke, CHD and total CV events. In a cohort of 5000 childhood-onset Type I diabetics followed for 28 years in Finland, 25% developed diabetic nephropathy. Of those, 7% had a stroke by the end of follow-up, compared to only 1.0% in those without diabetic nephropathy. These patients were young, under age 50 years, at end of follow-up.

Some of the effect of the increased risk of proteinuria may be due to hypertension, as other studies have shown that in Type I and II diabetic patients with diabetic neuropathy or microalbuminuria there is an increased prevalence of hypertension, (40% Type I, 50% Type 2). As the level of hypertension increases the urinary albumin excretion (UAE) rate also increases. This occurs at a greater rate in diabetic patients. A UAE level of 200 is considered diabetic nephropathy.

Blood pressure is known to be the strongest risk factor for stroke, with the risk of stroke increasing with increasing levels of blood pressure. This is not due solely to hypertension, though, as many patients with hypertension also have diabetes. A study from Finland of middle-aged persons showed that 12-16% with hypertension treated with drugs also had diabetes, compared to 2-4% among normotensives. Some of the effect may be due to diabetes, not just hypertension.

The Syst-Eur study of isolated systolic hypertension found that active treatment of hypertension in diabetic patients reduced total mortality by 65%, CV mortality by 76%, and all CV endpoints by 70% and all stroke by 73%. Effective blood pressure control reduces the risk of stroke more than 70% in diabetic patients. In the non-diabetic patients, there was no significant decrease in the total mortality, but there was a 40% decrease in stroke incidence.

Diabetes and even IGT are important risk factors for stroke and will be more so in the future due to the increasing number of diabetics. However, clinical data shows this risk can be decreased with proper treatment of blood glucose and effective control of hypertension. Postprandial glucose is the most dangerous, not the fasting level, so at present no country is properly using the oral glucose tolerance test (OGTT). The recommendation in Finland is that OGTT screening should be done in persons with hypertension, a history of diabetes, and in women with previous gestational diabetes.