The prevalence of diabetes in acute ischemic stroke was 31.8% in this study, similar to several other prospective stroke registry data [3,15, 16]. Also, among the patients diagnosed with diabetes, nearly 25% were so-called unrecognized diabetes, newly diagnosed by using HbA1C. Therefore, it has been recommended to check HbA1C or oral glucose tolerance test (OGTT) in all acute ischemic stroke patients, regardless of their previous history [17]. Among the recommended tests, HbA1C has an advantage over OGTT in identifying pre-diabetes in patients with an acute stroke unaffected by the acute-phase reaction [6]. Using HbA1C, the prevalence of pre-diabetes was 29.8% in our study, similar to previous reports. In this study, the presence of diabetes had a predictable value for the long-term outcome at 1-year but not for short-term outcomes.
Several previous results demonstrated that the presence of diabetes was significantly associated with END and worse outcomes at 90 days after acute ischemic stroke [5, 18]. However, in the Bichat stroke registry [4], the presence of diabetes had no impact on END and functional outcomes at 90 days in AIS using IV thrombolysis. Also, in the Lausanne Stroke Registry [19], there was no difference in short-term functional outcomes between stroke patients with and without diabetes. Differences in the patient population, functional outcome scales and follow-up duration may account for the contradictory results.
In particular, after acute stroke, short-term outcomes are primarily influenced by the patient's age and severity of stroke [20]. In our study, there were no differences of the age and stroke severity among three groups, so it was unlikely that the presence of diabetes would affect prognosis from the beginning of the stroke. In contrast, cardiovascular risk factors such as diabetes significantly impacted outcomes at least more than 90 days after AIS. This late-appearing pattern is probably related to the progression of the atherosclerotic vascular disease over time, which is known to be more rapid among diabetes. Also, the vascular risk profile among people with diabetes is worse because of their higher number of comorbidities and chronic conditions that result in accelerated atherosclerosis [21, 22]. In the Danish national stroke registry, diabetes did not affect patients' mortality with first ischemic stroke until 30 days; however, it began to affect it after 90 days and increased it significantly in 1 year [21]. Therefore, it is reasonable to conclude that diabetes affects the mid-to-long-term prognosis of acute stroke patients than their short-term outcomes.
Many previous studies showed that initial transient hyperglycemia higher than 140 mg/dl is an important predictor for in-hospital death after acute ischemic stroke using IV thrombolysis [2, 23, 24]. Our results also reaffirmed this notion. Data regarding the impact of transitory hyperglycemia on stroke mortality for more extended periods of follow-up are contradictory. 90 days after stroke onset, the mortality rate became almost equal between patients with transitory hyperglycemia and diabetes. Their rates did not differ after 180 or 365 days. Furthermore, upon admission, transient hyperglycemia was not an independent risk factor for one-year mortality after stroke [2, 25]. These reports indirectly support the authors' findings that transient hyperglycemia only affects the short-term prognosis after acute ischemic stroke.
In this study, pre-diabetes had a significant impact on the occurrence of END and in-hospital death after AIS using IV thrombolysis. Although a few previous studies suggested an association between the presence of pre-diabetes after acute stroke and worse outcomes [8-10], their relationship remains controversial [11, 12].
Generally, patients with pre-diabetes have an increased risk of type 2 diabetes and higher risks of cardiovascular diseases, such as stroke and recurrent stroke. Pre-diabetes, an intermediate metabolic state between normal glucose metabolism and diabetes, indicates an increased risk of developing type 2 diabetes and cardiovascular diseases in the future. Therefore, pre-diabetes should not be considered as a distinct clinical entity. Instead, it should be regarded as a continuum of increasing levels that represent growing risks of developing diabetes. Therefore, it is not logical to suggest that pre-diabetes and diabetes affect acute stroke in different directions.
In this data, AIS patients with pre-diabetes had a significantly lower prevalence of previous using statin (19.3% vs. 30.0 %, p<0.05, data are not shown) before AIS than those with diabetes. Also, the rate of previous using antiplatelet agents was lower in the pre-diabetes population than in the diabetes population. These data suggest that the pre-diabetes patients in our study may have been alienated from appropriate medical measures despite their risk of stroke occurrence. For this reason, they could have suffered more severe damages in the acute phase after stroke. Also, the increased hypercoagulable state in pre-diabetes patients without any interventions might have been critical for their early vascular instability [26, 27], consequently leading to END after AIS using IV thrombolysis. However, unlike diabetic patients, treatments such as using antiplatelet drugs immediately after an acute stroke and administering high doses of statins can ensure the stability of the vascular condition and have little effect on long-term prognosis.
Since only IV thrombolytic patients have been screened among AIS patients, this study has the advantage of reducing biases such as the onset time of stroke or treatment options that may affect long-term prognosis.
In this study, we proved that the presence of diabetes is associated with a worse long-term outcome after acute ischemic stroke. Also, we suggested that the presence of pre-diabetes may have an impact on the short-term outcome. A plausible reason may be the lack of interest in identifying pre-diabetes in the pre-stroke stage. However, there is a limitation in the confirmation above because this study is not a detailed prospective study and involves a small number of patients. In a future research, a systematic review of the short-term and long-term prognosis for pre-diabetes in patients with acute ischemic stroke that uses a meticulous large-scale acute stroke registry will be needed.