• J Fuller, L K Stevens, N Chaturvedi, and J F Holloway.
• Department of Epidemiology and Public Health, EURODIAB, University College of London, 1-19 Torrington Place, London, UK, WC1E 6BT. lynda@public-health.ucl.ac.uk
• Cochrane Db Syst Rev. 2000 Jan 1; 1997 (2): CD002188CD002188.

ObjectivesTo assess the effect of intervention, both pharmacological and non-parmacological, to reduce blood pressure in people with diabetes mellitus on all cause mortality, specific causes of death, including cardiovascular disease, stroke, ischaemic heart disease and renal disease, morbidity associated with macro- and microvascular complications of diabetes mellitus and also side effects of the interventions and their influence on quality of life and well being.Search StrategyThe search strategy employed was to searching electronic databases such as EMBASE and MEDLINE for all trials of anti-hypertensive treatment in diabetes mellitus. As well as searching specialist journals in the fields of cardiovascular disease, stroke, hypertension and renal diease.Selection CriteriaAll trials were considered independently and then discussed by 2 reviewers to determine there eligibility for inclusion in the review. Their methodological quality was also assessed from details of the randomisation methods, blinding and whether the intention-to-treat method of analysis was used. Trials included in the review were all randomised contolled trials of the treatment for anti-hypertensive therapy for the specified endpoints which included subjects with diabetes mellitus.Data Collection And AnalysisData was sought on the number of patients with diabetes with each outcome measure by allocated treatment group, either from previous publications or, if this was not possible, the raw data was obtained and analysed using the intention-to-treat method. If these data were not available the results from the 'Per Protocol' analysis were used. To compare the treatment effect of the intervention with that of placebo on all cause mortality and cardiaovascular mortality and morbidity, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for each trial and a meta analysis performed using Peto's ORs as the summary measure.Main ResultsThe initial search yielded 760 references, from which 23 appropriate trials were identified (3 for primary prevention and 20 for secondary prevention), and 15 of these trials had data available for analysis. For the primary prevention trials the summary ORs (95% CIs) for all cause mortality and CVD were 0.85 (0.62,1.17) and 0.64 (0.50,0.82) respectively. Of the seven trials for long-term secondary prevention (i.e. follow-up greater than one year), the summary OR (95% CI) for all cause mortality was 0.82 (0.69,0.99). Data on CVD mortality and morbidity was only available for 2 of these trials and the summary OR (95% CI) was 0.82 (0.60,1.13). There were five trials for short term secondary prevention trials (i.e. follow-up of less than 1 year) with data available for analysis. The summary ORs (95% CIs) for all cause mortality and CVD were 0.64 (0.50,0.83) and 0.68 (0.43,1.05) respectively.Reviewer's ConclusionsPrimary intervention trials indicated a treatment benefit for CVD, but not for total mortality in people with diabetes. For both short- and long-term secondary prevention, the present meta-analysis indicated a benefit for total mortality in diabetic subjects. However lack of information on CVD outcomes probably reduced the power of the meta-analysis to detect any corresponding benefit for this end-point. This, along with the fact that all published data of randomised control trials of anti-hypertensive therapy in diabetes for all cause mortailty and CVD outcomes are taken from the hypertension trials not specific to diabetes, underlines the need for further high quality trials examining the effects of blood pressure lowering interventions in people with diabetes.

12 keywords...

hide…