Abstract
Life expectancy for a patient with type 2 diabetes remains substantially shorter than an equivalent individual without diabetes, largely because of a greater risk of cardiovascular disease. Diabetes is also associated with an increased incidence of many types of cancer, suggesting that malignancy may also contribute to higher rates of mortality. Hyperglycemia is one of the key risk factors for diabetes-associated macro- and microvascular disease, and as such, intensive glycemic control is associated with improved outcomes for patients, including a reduction in this risk of death from any cause, when initiated early in the disease course. Recent trials in patients with more advanced disease have failed to demonstrate a mortality benefit with intensive glycemic control, although this may reflect their short observation period. Intensive multifactorial therapy, including lifestyle intervention and control of hyperglycemia, hypertension, lipids, thrombosis, and microalbuminuria, is likely to be the best strategy against diabetes-associated macrovascular mortality. However, analysis of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial indicates that there may be a subpopulation of patients who are unable to achieve glycemic targets with intensive therapy and that aggressive intensification of treatment in this group may increase mortality risk. It remains to be determined whether the relationship between diabetes and malignancy is causal or whether they share common risk factors. Current recommendations for a healthy lifestyle based on good diet, physical exercise, and weight management in order to control diabetes-related complications are likely to apply in reducing the risk of many forms of cancer and should be advocated for all patients.
Introduction
The major factor in increased mortality associated with diabetes is the risk of CVD, which is the leading cause of death in this population. Analysis of the First National Health and Nutrition Examination Survey covering the period 1971–1993 revealed that more than two-thirds of deaths of people with diabetes were due to CVD. 5 The most recent update of the Tayside observational cohort, based in Scotland, UK, followed a defined group of 10,532 patients newly diagnosed with type 2 diabetes and 21,056 comparators without diabetes from 1993 to 2004. 8 Vascular disease was the most frequent cause of death in both groups, but the incidence of such disease was considerably greater in patients with diabetes versus those without (45% vs. 39%). The study controlled for levels of deprivation, which is positively associated with higher rates of diabetes and mortality. Diabetes was associated with a 50% increase in cardiovascular mortality and 30% increase in all-cause mortality. 8 These figures are lower than in previous observational trials, which is likely to be owing, in part, to the design of the analysis, but may also reflect improvements in care of diabetes over the last decade. There are additional data suggesting that more stringent control of vascular risk factors advocated by management guidelines may have improved life expectancy for individuals with diabetes in recent years. Data from a defined cohort of 973 patients with type 2 diabetes in The Netherlands covering the period 2001–2007 reported no reduction in life expectancy compared with the general population, based on Dutch national statistics, but only in the absence of existing CVD or albuminuria. 9 The authors asserted that a normal life expectancy may be achievable in the future for patients with type 2 diabetes.
However, questions remain, including the rationale for data suggesting that intensive glycemic control, which is one of the cornerstones of current management guidelines, may be associated with increased mortality. 10 The aim of this article is to overview the latest evidence on the influences of mortality in patients with type 2 diabetes.
The Role of Hyperglycemia in Diabetes-Associated CVD and Mortality
The Diabetes Control and Complications Trial (DCCT) was the first large-scale randomized, controlled trial to conclusively demonstrate that intensive glycemic control could reduce the risk of microvascular complications in patients with type 1 diabetes compared with conventional therapy. 11 Patients in the intensive treatment arm showed a dramatic reduction in the incidence of diabetes-associated microvascular complications. There was no significant difference in the risk of macrovascular disease, although intensive therapy did reduce the development of hypercholesterolemia by 34% relative to conventional therapy, indicating an improvement in patients' risk profiles.
The DCCT cohort was followed up on an annual basis for another 11 years after the end of intervention as part of the Epidemiology of Diabetes Interventions and Complications (EDIC) study. 12,13 At the end of the DCCT, mean A1c levels were approximately 2% lower in the intensive treatment arm versus the conventional arm (EDIC population: 7.2% vs. 9.1%; P < 0.001). 12 This difference in A1c narrowed over the following 8 years of follow-up, stabilizing at approximately 8% for both groups (P = 0.83). Despite this convergence, follow-up at 11 years post-closure of the DCCT (17 years of total follow-up) revealed a significantly lower risk of macrovascular complications with initial intensive versus conventional therapy. 13 A total of 144 cardiovascular events occurred in 83 patients over this period, with a rate of 0.38 events per 1,000 patient-years in the intensive arm and 0.80 events per 1,000 patient-years in the conventional treatment arm (P = 0.007). Overall, intensive treatment was associated with a 42% relative reduction in risk of a first cardiovascular event (P = 0.02) and a 57% reduction in risk of nonfatal myocardial infarction, stroke, or death from CVD (P = 0.02).
The United Kingdom Prospective Diabetes Study (UKPDS) is the equivalent landmark trial for patients with type 2 diabetes. A total of 4,209 patients with newly diagnosed type 2 diabetes were randomized to receive conventional treatment or intensive therapy with a sulfonylurea, insulin, or, for overweight patients, a metformin-based regimen. 14,15 In the analysis of patients treated with a sulfonylurea and/or insulin, median A1c levels over 10 years of treatment were significantly lower with intensive versus conventional therapy (7.0% vs. 7.9%). 14 This difference was associated with a significant reduction in the relative risk for intensively treated patients of 25% for any microvascular end point (P = 0.0099), including a 21% reduction in the incidence of retinopathy (P = 0.015) and a 33% reduction in microalbuminuria (P < 0.0001). The relative risk of nonfatal myocardial infarction was 21% lower with intensive therapy, albeit with borderline statistical significance (P = 0.057). There was no significant difference in mortality rates, although there was a trend in favor of the intensive group, with a 10% relative risk reduction for diabetes-related death and 6% for all-cause mortality.
An analysis of the UKPDS data from 10 years after the end of the interventional period (total follow-up >20 years) provided further insight. 16 In this analysis, the relative risk of nonfatal myocardial infarction in the intensive treatment group was 15% lower and statistically significant (P = 0.01). The relative risk of death related to diabetes was 17% lower with intensive therapy (P = 0.01), and the risk of all-cause mortality was 13% lower (P = 0.007). The benefit against microvascular complications was also maintained over the extended follow-up period. Outcomes for the 342 overweight patients treated with metformin-based intensive therapy followed a similar pattern and were even greater in magnitude. These benefits were achieved despite the fact that the difference in A1c levels apparent at the end of intervention was lost within 1 year and were comparable over the remaining period of follow-up. This is analogous with the experience with type 1 diabetes patients in the EDIC study and demonstrates again the “legacy effect,” where the benefits of glycemic control persist long after intervention. Moreover, this also suggests that longer periods of observation than the approximately 5-year follow-up of the recent trials described below may be needed to see benefits on cardiovascular end points.
Is There a Link Between Hypoglycemia and Mortality in Patients with Diabetes?
Following the findings of the original UKPDS, several large-scale prospective trials were initiated to evaluate the benefits of intensive glycemic control against macrovascular disease in higher-risk populations (Table 1).
All values are mean (SD) unless stated otherwise.
Median over duration of follow-up period.
Events with impaired consciousness per 100 patient-years.
ACCORD, Action to Control Cardiovascular Risk in Diabetes; ADVANCE, Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation; NR, not reported; UKPDS, United Kingdom Prospective Diabetes Study; VADT, Veterans Affairs Diabetes Trial.
In the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial, 10,251 patients with type 2 diabetes and at high risk of CVD were randomized to intensive therapy (A1c, <6.0%) or to standard therapy (A1c, 7.0–7.9%). 10 The ACCORD population had a mean age of 62 years and a mean duration of diabetes of 10 years, with 35% already treated with insulin at baseline. Median baseline A1c levels were 8.1% and stabilized on study at 6.4% and 7.5% after 1 year of intensive or conventional therapy, respectively. At the interim analysis point (mean follow-up, 3.5 years), comparison of the intensive and conventional arms also revealed a significantly higher rate of death from any cause (5.0% vs. 4.0%; P = 0.04) and death from CVD (2.6% vs. 1.8%; P = 0.02) with intensive therapy, and the monitoring board prematurely terminated intensive treatment and switched all patients to the control arm. This was the first demonstration in a large-scale clinical trial that linked very aggressive blood glucose control with an increased risk of mortality in type 2 diabetes and caused much consternation. However, evidence from two additional trials published in the same year failed to support this ACCORD finding.
The Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) study was a second large-scale randomized trial comparing the impact of intensive glucose control with standard therapy. 17 As in the ACCORD study, the ADVANCE population were relatively old (mean age, 66 years) and were at high risk of CVD. However, compared with the ACCORD cohort, patients in the ADVANCE study had a mean duration of diabetes that was 2 years shorter (mean, 8 years), lower baseline A1c (median, 7.2%), and almost no use of insulin at enrollment. At median follow-up of 5 years, median A1c levels were 6.4% and 7.0% in the intensive and standard treatment arms, respectively. Intensive treatment was associated with a significant 14% reduction in the risk of microvascular events (P = 0.01), but the rate of macrovascular events was comparable between the two groups (P = 0.32). In contrast to the ACCORD study, intensive therapy was not associated with increased cardiovascular or all-cause mortality.
A third trial, the Veterans Affairs Diabetes Trial (VADT), randomized 1791 military veterans with type 2 diabetes to an intensive glycemic control (A1c, <6.0%) or standard glycemic control regimen. 18 Other CVD risk factors were treated aggressively and equally in both groups. Enrolled patients were uncontrolled on insulin or maximal-dose oral agents and represented a poorly controlled population with median baseline A1c of 9.4%. Mean age was 60 years, and the duration of diabetes was longer than in the ACCORD and ADVANCE studies, at 11.5 years. Median A1c levels of 6.9% and 8.5% were achieved in the intensive and standard arms, respectively, within the first 3 months of the study and maintained for the duration of the median 5.6 years of follow-up. There was no significant difference between the two groups in any one of the macrovascular events that formed the composite primary outcome. There was also no significant difference in the rate of microvascular events between the two groups and no difference in the rate of death from any cause (P = 0.62).
Based on the outcomes of these three trials and the contrast with the UKPDS, it appears that glycemic control may be more effective when initiated early in the course of disease before the underlying mechanisms that mediate vascular complications are established. The benefits of intensive glycemic control are also likely to be more evident with longer follow-up of more than 10 years, given that the power of ACCORD, ADVANCE, and VADT was limited in this respect. Several meta-analyses of the major trials assessing glycemic control in type 2 diabetes have been performed with data included from the UKPDS, ACCORD, ADVANCE, and VADT studies and others. 19 –21 Based on similar data sources, the analyses are consistent in showing a significant reduction in CVD with intensive versus conventional therapy and, in particular, a reduction in myocardial infarction. In the analysis by Ray et al., 20 intensive glycemic control resulted in a 17% reduction in rates of nonfatal myocardial infarction and a 15% reduction in coronary heart disease events; these benefits were achieved with no significant increase in the rate of all-cause mortality.
Insight into the benefits of early initiation of intensive therapy on macrovascular outcomes and mortality is likely to come from the ongoing “Outcome Reduction with an Initial Glargine Intervention” (ORIGIN) trial. The study has enrolled 12,612 subjects with evidence of CVD and with either impaired fasting glucose, impaired glucose tolerance, or newly detected or established diabetes (on zero or one oral antidiabetes agent). 22 Using a 2 × 2 factorial design, patients were randomized to initiate insulin glargine targeting an fasting plasma glucose of ≤95 mg/dL (5.5 mmol/L) or standard glycemic care and to ω-3 fatty acid supplements or placebo. The primary end points are the rate of cardiovascular events and cardiovascular mortality, respectively. To date, the risk of hypoglycemia appears to be low; 23 efficacy data are expected next year.
Evidence in support of the ACCORD finding on mortality comes from a recent retrospective, registry-based cohort study, by Currie et al., 24 who utilized data from a UK-based primary care database derived from patient records. Eligible patients had a diagnosis of type 2 diabetes, a case history of at least 6 months, and a documented history specifying escalation of their diabetes treatment. The primary outcome measure was all-cause mortality assessed according to post-index A1c, which was defined as the mean of all observations recorded after the first prescription of intensified diabetes therapy. The analysis revealed a U-shaped relationship between mortality and post-index A1c levels, with an increased unadjusted mortality rate at the highest and lowest A1c levels. The lowest mortality rate occurred at an intermediate A1c level of approximately 7.5%. However, the design of this study has been criticized, with the principal problem being the definition of A1c as a fixed variable rather than one that can change over time. Critics have also pointed out that studies in other fields have noted spurious U-shaped relationships between various biomarkers and mortality due to an association between extreme levels at either end of the spectrum and underlying disease. 25 For example, a U-shaped relationship was observed for cholesterol levels and mortality, which was influenced by a minority of critically ill patients with very low cholesterol levels; yet, clinical use of statins to lower cholesterol levels is not associated with increased mortality risk.
Despite numerous post hoc analyses, the reason for the excess mortality in the intensive arm of the ACCORD trial remains elusive. Several explanations have been proposed, including an excessively rapid reduction in A1c, increased severe hypoglycemia, weight gain, and high doses of certain antidiabetes agents or combinations of agents. In examining the first of these hypotheses, an epidemiological analysis of the interim data revealed that a moderate or large reduction in A1c over the first year of treatment was associated with similar rates of mortality in both the intensive and conventional treatment arms, whereas higher mortality rates were seen with intensive treatment when little or no reduction in A1c occurred 26 (Fig. 1). This observation is at odds with the “rapid A1c reduction” theory. The glycemic measure most strongly associated with mortality in the analysis was average A1c over the study period: a 1% increase in average A1c was associated with a 20% increase in the risk of death. This leaves the question as to why there was an excess of deaths in the intensive treatment arm despite mean A1c levels being significantly lower throughout the observed period. This may be partly explained by the fact that the correlation between higher A1c and death was greater in the intensive versus the conventional treatment arm. Plotting the risk of all-cause mortality against average A1c showed a linear increase in risk from 6% to >9% with intensive treatment (Fig. 2). In contrast, the conventional therapy plot showed a U-shaped relationship and a markedly lower risk of death relative to the intensive arm at average A1c levels >7% (Fig. 2). Therefore, it seems that patients in both groups who failed to achieve a notable reduction in A1c were at greater risk of death than those with an average A1c below 7%, but this increased risk was exaggerated in the intensive treatment arm.
The Action to Control Cardiovascular Risk in Diabetes (ACCORD) study: adjusted mortality rates by treatment strategy. 26 Curves display all-cause mortality rates by treatment for the whole period of follow-up, over a range of decreases in A1c from baseline in the first year of treatment (as a percentage of A1c). The calculations used a Poisson regression model with data from selected characteristics of participants and sites at baseline, severe hypoglycemia and weight change as time-varying covariates, and the randomization assignments in other ACCORD trials, plus assignment to the standard or intensive glycemic treatment strategies. The bold solid line represents the intensive treatment group, the bold dashed line represents the standard group, and the finer lines represent the 95% confidence intervals for each group. Adapted from Riddle et al. 26 © 2010. Reproduced with permission of the American Diabetes Association.
The Action to Control Cardiovascular Risk in Diabetes (ACCORD) study: risk of all-cause mortality (adjusted log[hazard ratio]) by treatment strategy. 26 Spline curves displaying the risk of all-cause mortality with the two treatment strategies over the range of average A1c from 6% to 9% relative to standard therapy at an A1c of 6%. The curves represent the linear part of the proportional hazards models derived from values for intervals of average A1c. The bold solid line represents the intensive treatment strategy group, the bold dashed line represents the standard group, and the finer lines represent the 95% confidence intervals for each group. Adapted from Riddle et al. 26 © 2010. Reproduced with permission of the American Diabetes Association.
The rates of hypoglycemia requiring medical or any assistance were significantly higher with intensive versus conventional treatment in the ACCORD study, 10 leading to the proposal that hypoglycemia may be responsible for the excess mortality in the intensive arm. In both treatment arms, patients who experienced an episode of severe, symptomatic hypoglycemia had a greater risk of death compared with those patients who had not experienced hypoglycemia. 27 However, among all patients who experienced a severe, symptomatic hypoglycemia episode, the risk of death was lower for patients in the intensive rather than the standard arm. This suggests that increased mortality in the intensive treatment arm cannot be attributed only to hypoglycemia. In fact, across both treatment groups, a 1% increase in patients' average A1c over the study period was associated with a 76% and 30% increase in the risk of severe hypoglycemia for the intensive and conventional treatment groups, respectively. 28 Echoing the results of the mortality analysis, patients with the smallest reduction from baseline in A1c were at greatest risk of severe hypoglycemia. 28
Combining these two lines of evidence from the ACCORD trial indicates that there may be a subpopulation of vulnerable patients who are unable to reduce their A1c level below 7% and that intensive therapy in this subgroup can increase the risk of mortality. This suggests that physicians who encounter patients with persistently high A1c levels should be cautious in intensifying treatment. Further study is required in order to identify and characterize subgroups with differing mortality risk so that treatment can be tailored appropriately.
Multifactorial Therapy to Reduce the Risk of Diabetes-Associated Mortality
It is well recognized that hyperglycemia represents only one of several potential risk factors for CVD and that control of all these factors may be necessary for optimal management. The Steno-2 study was a single-center, randomized, open-label, parallel-group trial that evaluated multifactorial therapy in 160 individuals with type 2 diabetes (mean duration, 6 years) and persistent microalbumiuria. 29,30 Eighty patients were randomized to receive conventional treatment for multiple CVD risk factors according to the Danish standard of care at the time of the study initiation in 1992. 30 The remaining 80 patients were randomized to intensive multifactorial therapy comprising lifestyle intervention and polypharmacy, with specific targets for controlling hyperglycemia (A1c <6.5%), hypertension, lipids, thrombosis, and microalbuminuria. The behavioral, clinical, and biochemical characteristics of the two study groups were comparable at baseline but had diverged considerably in favor of the intensive group after a mean follow-up of 7.8 years. These differences were reflected in the risk of CVD, which, for patients who received intensive treatment, was more than 50% lower compared with conventional therapy (hazard ratio [HR], 0.47 [95% confidence intervals [CI]: 0.24–0.73]). The risks of nephropathy, retinopathy, and neuropathy were also significantly lower. Rates of hypoglycemia were comparable between the groups.
A follow-up analysis of the Steno-2 study was conducted 5.5 years after the end of intervention, by which time the characteristics of the two groups had converged, primarily because of intensified therapy in the original conventional therapy group. Over the entire 13.3-year period of follow-up, 24 deaths (30%) in the intensive group and 40 deaths (50%) in the conventional group had occurred, which translated into a 46% reduction in mortality with intensive therapy (HR, 0.54 [95% CI: 0.32–0.89]; P = 0.02). 29 Intensive therapy was also associated with a lower risk of death from cardiovascular causes (HR, 0.43 [95% CI: 0.19–0.94]; P = 0.04) and of cardiovascular events (HR, 0.41 [95% CI: 0.25–0.67]; P < 0.001). 29
The ACCORD study also evaluated the impact of intensive control of hypertension or dyslipidemia on cardiovascular outcomes and mortality. Of the 10,251 patients randomized to intensive or standard glycemic control, 5,518 were also randomly assigned (in a 2 × 2 factorial design) to either simvastatin plus fenofibrate or simvastatin plus placebo (the ACCORD lipid trial), 31 and the remaining 4,733 participants were also randomly assigned to either intensive (<120 mm Hg) or standard (<140 mm Hg) blood pressure control (the ACCORD blood pressure trial). 32 As in the original study, the primary outcome for both subtrials was the composite of first occurrence of nonfatal myocardial infarction, nonfatal stroke, or death from cardiovascular causes. With a mean follow-up of 4.7 years, both studies failed to show any significant benefit in either the primary or secondary outcomes, with the exception of a reduced risk of any and nonfatal stroke with intensive versus standard blood pressure control. 31,32 Mortality rates were comparable between the two groups in both trials. As in the case of glycemic control, the benefits of controlling hypertension and dyslipidemia may take several years to become apparent or may be less evident in patients at higher risk who are likely to have more established risk factors that are not easily reversible.
The Potential Role of Cancer in the Mortality of Patients with Type 2 Diabetes
Epidemiological evidence suggests that diabetes may be associated with an increased incidence of many cancers, indicating that death due to malignancy could be an additional reason for higher rates of mortality with diabetes. Diabetes is a recognized risk factor for a range of cancers, including bladder, breast, colorectal, endometrial, liver, and pancreatic cancer and non-Hodgkin's lymphoma. 33 However, it is not known whether these associations are causal or whether they are due to the fact that both diabetes and cancer share several common risk factors, including obesity, low physical activity, poor diet, high alcohol consumption, and smoking.
Mortality rates in obesity, diabetes, and cancer populations are high, with both obesity and type 2 diabetes being independently associated with an increased risk of cancer-related mortality. 34 –41 Evidence from a large, prospective U.S. cohort shows diabetes is an independent predictor of mortality from cancer of the colon, pancreas, and breast in women and of the liver and bladder in men. 34 There is also evidence to suggest that diabetes may significantly increase mortality in patients with cancer. Meta-analyses of data from selected cancer sites demonstrate that preexisting diabetes increases the risk of long-term mortality in patients with endometrial, breast, and colorectal cancer. 37,42,43 Mortality risk also increases in postoperative cancer patients with diabetes. 44
Several plausible biological mechanisms have been put forward for the link between diabetes and cancer, including the effects of hyperglycemia, hyperinsulinemia, and inflammation. The Warburg effect is based on the observation that cancer cells produce energy primarily by glycolysis, which creates a high requirement for glucose. Thus, hyperglycemia may facilitate the proliferation of malignant cells. Elevated fasting serum glucose levels have been shown to be an independent risk factor for certain cancers, and the risk increases with increased glucose levels, as does cancer-related mortality. 45 A prospective cohort study of more than 60,000 Swedish individuals identified a significant increase in cancer risk for patients in the highest quartile for fasting and postload plasma glucose. 46 However, there is no conclusive clinical evidence to suggest that good glycemic control reduces the risk of cancer. A recent meta-analysis of large-scale randomized controlled trials in patients with type 2 diabetes, including the UKPDS, ACCORD, ADVANCE, and VADT, failed to find a relationship between reduced blood glucose levels and improvements in the risk of cancer incidence or mortality due to cancer. 47 There is a suggestion from a meta-analysis of epidemiological trials that metformin-treated patients may have a lower risk of some malignancies versus other treatment types, although the risk of selection bias in nonrandomized trials means that prospective data from randomized controlled trials are required to confirm this observation. 48 There has also been a suggested link between the use of insulin analogs and malignancy in epidemiological studies, although other studies show no such evidence. 49 –53 The ongoing ORIGIN study, with more than 12,000 patients, will be the largest prospective study to date examining malignancy risk with insulin analog therapy as a secondary end point.
Hyperinsulinemia is a second possible mechanism by which diabetes may influence cancer risk. The human insulin receptor (IR) is expressed as two isoforms, IR-A (short form) and IR-B (long form). 54,55 There is considerable evidence that IR-A, which binds to insulin and insulin-like growth factor-2, may play a critical role in the development of breast cancer 56 and other malignancies. 57,58 Activation of both the IR and insulin-like growth factor-1 receptors also results in up-regulation of a variety of intracellular processes, in part through shared intracellular signaling pathways for metabolic and mitogenic processes, the latter of which can result in the proliferation of cancers. 59
Further study is required in order to determine more precise mechanisms for the increased cancer risk in patients with diabetes before optimal strategies for management and prevention of the risk can be formulated. Current recommendations for a healthy lifestyle based on good diet, physical exercise, and weight management in order to control diabetes-related complications are likely to apply in reducing the risk of many forms of cancer and should be implemented for all patients.
Conclusions
Diabetes remains a condition associated with an increased risk of mortality with macrovascular disease responsible for a large proportion of the excess risk. The follow-up of the UKPDS 10 years post-cessation of intervention suggests that establishing glycemic control early in the disease course with intensive treatment can significantly improve outcomes, including a reduced risk of all-cause mortality. 14 Control of additional risk factors, including lipid profiles and hypertension, with multifactorial treatment regimens is likely to confer additional benefits. 29 Acting early after diagnosis is important based on the fact that the improved outcomes persist many years after initial intervention and even if glycemic control is subsequently lost. 12,13,16 However, the question of whether an intensive treatment strategy reduces the risk of macrovascular disease and mortality in populations with more advanced disease remains to be answered. The large-scale trials conducted in high-risk patients to date have not been of sufficient duration to address this question fully. Further study is required in this area, particularly given that the reasons for the excess mortality in the intensive arm of the ACCORD trial remain a puzzle. There is a possibility that there is a subset of patients with type 2 diabetes at high risk of vascular disease for whom intensive treatment may be life-threatening, and further study is needed to confirm or discount this hypothesis. The link among diabetes, cancer, and mortality is another area of increasing interest. Evidence to date is based largely on epidemiological trials, and well-controlled, prospective trials are needed. Strategies to tackle mortality due to vascular disease are now well established, and further study is required to identify mechanisms, risk factors, and strategies to combat the risk of malignancy in diabetes populations. Considerable progress has been made in mitigating the increased risk of mortality associated with type 2 diabetes, but we have further to go if we are to achieve the ultimate aim of a normal life expectancy.
Footnotes
Acknowledgments
Editorial support for this article was provided by Huw Jones, Ph.D., of Medicus International and was funded by sanofi-aventis.
Author Disclosure Statement
G.D. reports receiving consulting fees from sanofi-aventis and Merck, grant support from Medtronic, and honoraria from sanofi-aventis and Merck.