Objective

Diabetes mellitus is a well-established risk factor for atherosclerotic disease, but its role in the occurrence of venous thromboembolism (VTE) has not been elucidated. We conducted a meta-analysis of published cohort and case-control studies to assess whether diabetes mellitus is a risk factor for VTE.

Research design and methods

We systematically searched MEDLINE and EMBASE for case-control and prospective cohort studies assessing association between the risk of venous thromboembolism and diabetes. Odds ratios (OR) from case-control studies were combined while for prospective studies hazard ratios (HR) were combined. Models with random effects were used. Meta-analyses were conducted separately for raw and adjusted measures of association.

Results

24 studies were identified including 10 cohort studies (274,501 patients) and 14 case-control studies (1,157,086 patients). Meta-analysis of the prospective cohort studies demonstrated a significant association between diabetes and VTE (HR 1.60; 95% CI 1.35 to 1.89). This association was no longer present after analysis of multi-adjusted HRs (HR 1.10; 95% CI 0.77 to 1.56). Meta-analysis of case-control studies showed a significant association between diabetes and VTE (OR 1.57; 95%CI 1.17 to 2.12), but this association was no longer present when adjusted ORs were used (OR 1.18; 95%CI 0.89 to 1.56).

Conclusions

The increased risk of VTE associated with diabetes mainly results from confounders rather than an intrinsic effect of diabetes on venous thrombotic risk. Therefore, no specific recommendations should apply for the management of diabetic patients at risk for VTE.