Diabetes common in stroke, tied to poorer outcomes
Diabetes is relatively common among stroke patients and appears to be linked to poorer outcomes, according to a recent meta-analysis.
Accessing the databases of Ovid Medline and Embase, researchers identified studies that described the prevalence and influence on outcomes of diabetes or acute hyperglycaemia in stroke patients. A total of 66 studies were eligible for inclusion, corresponding to a total combined sample size of 522,645 patients.
Pooled analysis of 39 studies (n=359,783) showed that the prevalence rate of diabetes in stroke inpatients was 28 percent, though heterogeneity among the studies was significant (p=0.017). This was higher among those with ischaemic stroke than haemorrhagic stroke (33 percent vs 26 percent).
When glycated haemoglobin (HbA1c) was used as the sole diagnostic criteria for diabetes, the prevalence rate in all stroke types further shot up to 37 percent.
There were large variations in the documented prevalence of diabetes. The rates of known diabetes ranged from 8.2 percent to 56.2 percent, while that of unrecognized diabetes ranged from 5 percent to 33 percent. In studies that used HbA1c and a history of diabetes as diagnostic criteria, the resulting prevalence rates fluctuated between 24.7 percent and 56.2 percent.
In terms of outcomes after stroke, six studies confirmed a link between diabetes and poorer neurological outcomes, while four found no significant effect. There were also potential correlations of diabetes with longer in-hospital lengths of stay, higher readmission rates and greater risks of stroke recurrence.