Review: diabetes is not a coronary disease equivalent
Do patients with diabetes and no history of myocardial infarction (MI) have similar risk of coronary artery disease (CAD) events as patients with previous MI and no diabetes?
Included studies compared risk for CAD events in patients with diabetes and no MI and patients with previous MI and no diabetes. Outcomes were fatal or non-fatal MI.
Medline and EMBASE/Excerpta Medica (to Apr 2007); Cochrane Central Register of Controlled Trials; and references were searched for cohort studies involving ⩾100 patients with follow-up ⩾4 years. 13 articles reporting on 14 cohorts (n = 47 361, age 25–84 y, 0–100% men) met the selection criteria. Mean duration of follow-up was 13 years (range 5 to 25 y). Patients were enrolled at either a cross-sectional point in time (12 cohorts) or at diagnosis of diabetes or first MI (2 cohorts).
Compared with patients with previous MI and no diabetes, those with diabetes and no MI had lower risk of a CAD event (table).
Patients with diabetes and no history of myocardial infarction (MI) have a lower risk of coronary artery disease events than patients with previous MI and no diabetes.
Abstract and commentary also appear in ACP Journal Club.
Bulugahapitiya U, Siyambalapitiya S, Sithole J, et al. Is diabetes a coronary risk equivalent? Systematic review and meta-analysis. Diabet Med 2009;26:142–8.
Clinical impact ratings: Cardiology 7/7; IM/Ambulatory care 6/7; Endocrine 5/7
In recent years, evidence from large cohort studies, including the Finnish study by Haffner et al, has led to the concept that diabetes mellitus (DM) is a coronary risk equivalent.1 2 In trials of acute coronary syndromes, patients with DM have increased risk of subsequent cardiovascular events. Therefore, experts have argued that patients with DM and no coronary events should be treated in a similar fashion to their counterparts with CAD.
To evaluate this concept, Bulugahapitiya et al evaluated 14 cohorts in a systematic overview and found that patients with DM and no history of MI had lower risk of CAD events than patients without DM and a history of MI. The discrepancies between this meta-analysis and the Haffner study can be explained by the fact that diabetes, as a cardiovascular process, is complex.
Simply, not all primary prevention patients with DM are the same. The cardiovascular prognosis for type 2 DM depends on several variables, including age, duration of DM, degree of optimisation of serum lipid concentrations and blood pressure, and presence of other comorbid conditions, such as nephropathy. Because type 2 DM is often accompanied by ⩾1 other traditional risk factor, some rationale exists for treating patients with DM aggressively. Based on national guidelines over the past decade, we have witnessed the advent of statin therapy for all patients with type 2 DM regardless of low-density lipoprotein cholesterol concentrations, an aggressive blood pressure target of 130/80 mm Hg, and aspirin for all patients who reach middle age. Now is the time to reflect on the need to individualise therapy by weighing the risks and benefits of these therapies with our patients, while considering their estimated risk of coronary events, rather than assuming that all share the same high risk.
Given that patients with type 2 DM are at higher risk of CAD than patients without DM, national guidelines from the American Diabetes Association (ADA) and American Heart Association (AHA) have recommended optimisation of medical risk factors in all such patients.34 The 2 societies differed on whether to screen asymptomatic patients with DM for CAD, with the ADA recommending routine stress testing to detect coronary ischaemia and the AHA taking a more conservative approach. This debate served as the primary question in the DIAD trial by Young et al. DIAD randomised asymptomatic patients with type 2 DM to a screening nuclear imaging stress test or usual care.
This landmark trial enrolled >1000 patients and achieved excellent control of risk factors throughout the 5-year follow-up. Overall, the primary outcome of cardiac death or non-fatal MI did not differ between the screened and usual-care groups. The low event rate can be attributed largely to the aggressive medical care achieved equally in both groups. For example, 67% of patients were receiving a statin drug at 5 years.
DIAD illustrates the relatively low yield of screening to identify a high-risk cohort: In the screening group, only 6% of patients had a moderate-to-large perfusion defect. This group was at higher risk of CAD events, underscoring that more effective risk stratification short of a stress test may be desirable if a patient falls within this group. The sample size of DIAD was too small to determine whether treating these high-risk patients conservatively is completely safe, but recent data from the BARI 2D trial showed that when these patients were referred for coronary revascularisation, there was no difference in long-term mortality compared with aggressive medical therapy alone.5
Although the Bulugahapitiya meta-analysis seems to settle the debate about whether type 2 DM is a coronary risk equivalent, it may have little effect on clinical practice. Patients with type 2 DM do not require routine screening for coronary ischaemia, provided that they are treated with the current standard of care, an aggressive medical strategy.