File(s) under permanent embargo
Risk of all-cause mortality and vascular events in women versus men with type 1 diabetes: A systematic review and meta-analysis
journal contributionposted on 2015-03-01, 00:00 authored by Rachel HuxleyRachel Huxley, S A E Peters, G D Mishra, M Woodward
© 2015 Elsevier Ltd. Background: Studies have suggested sex differences in the mortality rate associated with type 1 diabetes. We did a meta-analysis to provide reliable estimates of any sex differences in the effect of type 1 diabetes on risk of all-cause mortality and cause-specific outcomes. Methods: We systematically searched PubMed for studies published between Jan 1, 1966, and Nov 26, 2014. Selected studies reported sex-specific estimates of the standardised mortality ratio (SMR) or hazard ratios associated with type 1 diabetes, either for all-cause mortality or cause-specific outcomes. We used random effects meta-analyses with inverse variance weighting to obtain sex-specific SMRs and their pooled ratio (women to men) for all-cause mortality, for mortality from cardiovascular disease, renal disease, cancer, the combined outcome of accident and suicide, and from incident coronary heart disease and stroke associated with type 1 diabetes. Findings: Data from 26 studies including 214 114 individuals and 15 273 events were included. The pooled women-to-men ratio of the SMR for all-cause mortality was 1·37 (95% CI 1·21-1·56), for incident stroke 1·37 (1·03-1·81), for fatal renal disease 1·44 (1·02-2·05), and for fatal cardiovascular diseases 1·86 (1·62-2·15). For incident coronary heart disease the sex difference was more extreme; the pooled women-to-men ratio of the SMR was 2·54 (95% CI 1·80-3·60). No evidence suggested a sex difference for mortality associated with type 1 diabetes from cancer, or accident and suicide. Interpretation: Women with type 1 diabetes have a roughly 40% greater excess risk of all-cause mortality, and twice the excess risk of fatal and nonfatal vascular events, compared with men with type 1 diabetes. Funding: None.