Patients with diabetes are at high risk of coronary heart disease. Leading organisations have recommended that all diabetic patients should be treated as aggressively as patients with established coronary heart disease.1 Randomised trials have shown the efficacy of reducing low density lipoprotein concentrations in patients without coronary heart disease. Large trials and meta-analyses of such trials would be expected to provide information on diabetic patients. We therefore systematically examined the available data on lowering low density lipoprotein concentrations in diabetic patients without coronary heart disease.
Methods and results
The review protocol is available from the authors. We searched Medline and eight other electronic databases (including five clinical trials databases) and proceedings from pertinent scientific meetings. We attempted to contact the authors of trials reporting incomplete data but received no responses. We reviewed the bibliographies of all retrieved publications.
Eligible trials randomised patients to lipid lowering interventions; included patients without coronary heart disease; and measured myocardial infarction, death from coronary heart disease, or all cause mortality. Eligible meta-analyses pooled data from similar trials. We excluded studies available only as abstracts. There were no language exclusions. We also included the Medical Research Council/British Heart Foundation heart protection study, which was published after our search.2 A list of included trials and meta-analyses is available on bmj.com
The 14 eligible trials randomised 132 977 patients without coronary heart disease, and diabetes status was stated for 1799 patients (1.3%). Three trials provided clinical endpoints for 454 diabetic patients. In addition, the heart protection study randomised 3982 diabetic patients without coronary heart disease (table).
We found 13 meta-analyses that included up to 38 trials and 146 854 patients. None presented data for diabetes subgroups. One meta-analysis postulated that diabetes might account for differences between trials, but incomplete reporting in the trials limited the analysis.
Comment
Inclusion and reporting biases in randomised controlled trials and meta-analyses limited our assessment of the efficacy of lowering low density lipoprotein concentration in diabetic patients without coronary heart disease. Most trials of lipid lowering interventions for primary prevention of coronary heart disease excluded diabetic patients by varied and ambiguous criteria. Consequently, these trials included few patients with diabetes. Those who were included were poorly characterised in terms of type and duration of diabetes, severity of complications, and metabolic control. It is therefore unclear whether the diabetes subgroups represent the general diabetic population.
The meta-analyses were also affected by the limitations described above. In addition, outcome was reported for only one third of diabetic participants. Consequently, these meta-analyses cannot overcome the biases against diabetes in the original trials.
Current recommendations to manage dyslipidaemia in diabetic patients are based on observational evidence and expert judgment. The heart protection study showed that simvastatin significantly reduced the risk of major vascular events for diabetic patients without coronary heart disease at any initial low density lipoprotein concentration. It remains unclear whether the benefits of statins are mediated by lowering low density lipoprotein concentrations, whether goals of treatment should be expressed as low density lipoprotein concentrations, and whether a fixed dose of statin, increasing doses of statin, or multiple drugs can be used to achieve these goals with acceptable safety. Recommendations from policymakers and experts should reflect this uncertainty.
Supplementary Material
Table.
Trial (year)*
|
Mean or median follow up (years)
|
No of patients randomised
|
Diabetic patients excluded
|
Diabetes exclusion criteria
|
Exclusion rationale
|
Baseline diabetes status published
|
No of diabetic patients randomised
|
Risk ratio (95%CI) for first coronary heart disease event
|
|
---|---|---|---|---|---|---|---|---|---|
All patients
|
Diabetic patients
|
||||||||
Diet | |||||||||
MRFIT (1982)w1 | 7 | 12 866 | Some | “Requiring medication” | None | No | Unknown | 0.92 (0.80 to 1.05) | NR |
WHO Multifactor (1983)w2 | 6 | 49 781 | No | NA | NA | No | Unknown | 1.08 (0.98 to 1.19) | NR |
Gothenberg (1986)w3 | 11.8 | 30 000 | No | NA | NA | No | Unknown | 0.99 (0.91 to 1.07) | NR |
Resins | |||||||||
LRC-CPPT (1984)w4 | 7.4 | 3 806 | All | Medical history or fasting blood glucose>130 mmol/l | None | NA | NA | 0.83 (0.67 to 1.01) | NA |
Fibrates | |||||||||
VA Cooperative (1973)w5 | 4.5 | 532 | Some | “Severe” | None | Yes | 47 | 0.90 (0.41 to 1.97) | NR |
WHO Cooperative (1979)w6 | 5.3 | 15 745 | Some | “Requiring drug treatment” | None | No | Unknown | 0.83 (0.68 to 1.01) | NR |
Helsinki Heart Study (1987)w7 | 5 | 4 081 | Some | All but “mild” or insulin use | None | Yes | 135 | 0.66 (0.47 to 0.92) | 0.32 (0.08 to 1.27) |
SENDCAP (1998)w8 | 3 | 164 | No | NA | NA | Yes | 164 | NA | 0.32 (0.13 to 0.79) |
DAIS (2001)w9 | 3 | 200 | No | NA | NA | Yes | 200 | NR | NR |
Statin trials | |||||||||
Multination Study Group (1993)w10 | 0.5 | 1 062 | All | “Significant endocrine disease” | None | NA | NA | NR | NA |
Oxford (1994)w11 | 3.4 | 621 | Some | Untreated | None | Yes | 20 | NR | NR |
WOSCOPS (1995)w12 | 4.9 | 6 595 | No | NA | NA | Yes | 1 037 | 0.70 (0.58 to 0.84) | NR |
ACAPSw13 | 2.8 | 919 | No | NA | NA | Yes | 21 | 0.36 (0.13 to 0.94) | NR |
AFCAPS/ TexCAPS (1998) w14 |
5.2 | 6 605 | Some | Uncontrolled or insulin use | None | Yes | 155 | 0.63 (0.50 to 0.79) | 0.56 (0.18 to 1.79) |
Heart Protection Study (2002)2 | 5 | 7 150 | No | NA | NA | Yes | 3 982 | 0.77 (0.70 to 0.85) | 0.74 (0.64 to 0.86) |
NA=not applicable, NR=data not reported.
Names of trials are given in full in the references, which are available on bmj.com
Footnotes
Funding: American Diabetes Association and Mayo Foundation. The guarantor accepts full responsibility for the conduct of the study, had access to the data, and controlled the decision to publish.
Competing interests: None declared.
Ethical approval: The study did not involve human subjects or patients' records. The institutional review board of the Mayo Clinic approved the study.
The studies included in the review are listed on bmj.com
References
- 1.Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. National cholesterol education program: executive summary of the third report of the Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (adult treatment panel III) JAMA. 2001;285:2486–2497. doi: 10.1001/jama.285.19.2486. [DOI] [PubMed] [Google Scholar]
- 2.Medical Research Council/British Heart Foundation Heart Protection Study Collaborative Group. MRC/BHF heart protection study of cholesterol lowering with simvastatin in 20 536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002;360:7–22. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.