An Observational Study of Glycemic Status in New Onset Acute Stroke and Their Clinical Outcome in a Tertiary Care Teaching Hospital

Sample selection: 158 patients were selected by using randomized table method. Written informed and valid consents were taken from the patients after providing adequate information and answering their question and queries in detail.

Method: All patients were studied for clinical history, blood pressure, blood sugar, urea, creatinine, electrolytes, haemoglobin, total cell count, differential count, urine sugar, albumin, electrocardiogram, chest X-ray, CT scan brain (if require), and 2D-ECHO. The severity of stroke for each patient was calculated based on NIH stroke scale, NIHSS which takes the following clinical findings into account and each criteria awarded specific points (14,15).

The points were added, with a maximum of thirty points.

Once clinical diagnosis of acute stroke was made venous blood sample was taken, with in twenty-four hours of onset of symptoms, and sent to laboratory for glucose estimation.

In patients with blood sugar more than 6.1 mmol/L (110 mg/dl), Haemoglobin A1c (HbA1c) was performed, and diabetic status noted.

(HbA1c is structurally similar to haemoglobin A except for the addition of glucose Group to the terminal amino acid of the beta chain of the haemoglobin Molecule glycosylation).

Therefore, HbA1c is a function of the exposure of the red blood cells to glucose. Since the glucose linkage to haemoglobin is relatively stable, Hemoglobin A1c accumulates through out of the life span of erythrocyte and its concentration reflects the integrated blood glucose concentration over a period approximating to the half-life of erythrocytes, i.e. six to eight weeks. Measurement of HbA1c helps to monitor the overall degree of diabetic control achieved. The normal range of Hemoglobin A1c is 3.8% to 6.4%.

The patients can be classified into four groups (16):

Blood sugar < 6.1 mmol/l [110mg/dl]: Nondiabetic/euglycemic.

Then computerized tomography (CT) of the brain was performed in all patients:

CT scan immediately done to rule out haemorrhagic stroke.

Patient was given corticosteroid according to protocol with vasogenic edema hemodynamically stable patients. Injection dexamethasone 8mg was given thrice daily and tappered accordingly after improvement on repeat CT scan.

The patients were assessed on the day of admission NIHSS score and the day 15 NIHSS Score outcome observed in the form of death, poor, moderate, and good prognosis.

Patients who were unable to return to any form of work, persistent disability, need for residential placement, dependent in activities of daily living, and stable deficit with no recovery were classified as those with poor outcome.

Patient whose symptoms improved, who were independent in attending day to day activities, improvement in motor function and aphasia, and no persistent disability were grouped as patients with good outcome.

Patients who fared in between these two groups were grouped as those with moderate outcome.

Statistical Analysis: The collected data entered into the Excel spread sheet and transferred to the SPSS version 23 for statistical analyses (IBM, Chicago, Illinois). Chi-square test was used as test of significance for categorical data. Unpaired t test was used for continuous data. P value less than 0.05 was considered as statistically significant.

Major risk factor was hypertension, i.e. 65.8% (104 cases), followed by smoking 39.8 % (63 cases), Diabetes in 27.8 % (44 cases) and Alcohol 25.9% (41 cases). Among major risk factor in males, smoking was found in 100%, alcohol was found in 96.1%, hypercholesterolemia was found in 71.4%, and hypertension was 63.6%. Whereas in females Diabetes was found in 39.3% and hypertension was found in 36.4% (Table 1).

Among neurological symptoms, maximum cases were presented with right hemiplegia in 55% (n=87), followed by loss of consciousness in 50.6% (n=80), left hemiplegia in 38.6% (n=61), aphasia in 37.9 %(n=60), and bladder & bowel involvement in 26.5%(n=42). Male patients were commonly presented with cerebellar symptoms, right hemiplegia, left hemiplegia, and loss of consciousness. Whereas faciobrachial monoplegia was predominantly seen in female patients (Table 1).

44.3% cases were euglycemic, 27.8% cases were stress hyperglycemic, 15.9% cases were known diabetic, and 12% cases were newly diagnosed diabetics.

Maximum NIHSS score was found in stress hyperglycemic patients (19.4%), followed by known diabetics (17.3%), newly detected diabetics (16.3%), and euglycemics (9.5%). This association was statistically significant (P<0.05) (Table 2).

Maximum proportion (81.8%) of ischaemic stroke cases were found with stress hyperglycemia compared to other glycemic conditions but in hemorrhagic stroke maximum proportion (47.3%) were newly diagnosed diabetics. This association was statistically significant (P<0.05) (Table 3).

Maximum proportion of large size of lesion (47.4%) was found in newly diagnosed diabetes followed by stress hyperglycemia (45.55). Medium size of the lesions were found more in newly diagnosed diabetes (52.6%) followed by known diabetes cases (52%). This association was statistically significant (P<0.05) (Table 4).

Among the 158 study subjects, hemorrhagic stroke was seen in 43 (27.2%) patients and ischaemic stroke was seen in 115 (72.8%) patients (Table 5).

As NIHSS score increases, the severity of the disease was also increased. Maximum percentage (50%) of deaths were seen in newly diagnosed diabetes cases compared to other groups. Maximum cases of good outcome belonged to euglycemic group (65.9%). This association was statistically significant (P<0.05); (Figure 1).

Fig. 1

All the good outcome cases in hemorrhagic stroke belonged to euglycemia group and in ischemia group also maximum cases of good outcome belonged to euglycemia. Maximum proportion of death cases belong to stress hyperglycemia in hemorrhagic stroke group (5/8=62.5%) (Table 6).

Maximum proportion of death cases in ischemic group belongs to newly diagnosed diabetes cases (8/9=88.8%).

More proportion of deaths were found in blood sugar (<110 mg/dl) and blood sugar (>199 mg/dl) such as 41.2% and 53.4%. Mortality was high in hyperglycemic cases in hemorrhagic stroke cases. In ishchaemic stroke maximum mortality was found with hyperglycemic patients (88.9%) (Table 7).

Among the 158 patients, 104(65.8%) cases were hypertensive with Male preponderance of in 66(63.6%) than females in 8(36.36%). More than half of diabetic patients were males in 27(60.7%) when compared to females in 17(39.3%). The risk of diabetes mellitus was 27.8%. More than two third of hypercholesterolemia patients were males in 16(71.4%) than in females 6(28.56%).

O' Donnell MJ et al² showed concordance with our study. This study showed patients with risk factors of hypertension (99%), diabetes mellitus (5%), alcohol intake (3.8%), cardiac causes (6.7%). Therefore, showing hypertension, smoking, diabetes, alcohol intake were significant risk factors for hemorrhagic stroke.

Among the 158 patients in our study group 70 patients had normal blood glucose values. Raised blood glucose levels on admission were noticed in 19(12%) newly diagnosed diabetes, 25(15.9%) known diabetes, and 44(27.8%) stress hyperglycemia patients. Other study Singh KG et al showed preponderance of known diabetes 14(28%) in hyperglycemic group.

In ischemic stroke group stress hyperglycemia amounted to one third of the patients and one fifth in hemorrhagic group. The present study shows 15.9% (25/158) prevalence of known diabetics presenting with acute stroke; whereas the prevalence of stroke in known diabetics was 8.5%. But in a study, Kyadav K et al (20) known diabetics were high compared to the present study (24%).

The study also shows a high prevalence of newly diagnosed diabetics (12%) in patients presenting with acute stroke. In another study, the incident of stroke was 16% and 12% in known diabetics and newly diagnosed diabetics (10), which is similar to the present study. These are in agreement with the observations in other series (23). Previous study found a range of prevalence of undiagnosed diabetes in acute stroke population from 6% to 42% (24).

Severity of stroke was assessed with NIH Stroke scaling system (NIHSS) score. On the admission day hyperglycemic patients had a higher score when compared to euglycemic patients (17.27 vs. 9.5) which was statistically significant with p = 0.001. Among the admission day hyperglycemic patient‘s stress hyperglycemics had higher NIHSS score (19.4%). As score increases severity also increases.

Similar concordance was seen in Al-Weshshy A et al (25) showing NIHSS score high in hyperglycemia patients (14.9±5.9 vs 7.8±3.5, p=0.000). 38.1% of patients with stress hyperglycemia had high NIHSS score versus 10% diabetics and 5% control. 30 days mortality higher in stress hyperglycemia compared to diabetics.

The size of the lesion was analyzed with the help of CT scan brain. Most of the euglycemic 46(65.7%) patients had small sized infarcts and hemorrhage whereas majority of the admission day hyperglycemic patients had large sized lesion with edema and midline shift. Among the hyperglycemic patients large sized lesion observed in newly diagnosed diabetes in 9(47.7%) and stress hyperglycaemia group in 20(45.5%). Similar result found in Singh KG et al (21) showing large sized lesions in stress hyperglycemia 5(100%) and also in newly diagnosed diabetes 11(100%).

Among the euglycemic group three fourth of the patients had ischemic stroke and one fourth had hemorrhagic stroke. Among patients with admission day hyperglycemia one third of the patients had hemorrhagic stroke. In newly detected diabetic patients, half of them had hemorrhagic stroke. Our study shows an increasing incidence of hemorrhagic stroke among diabetes patients.

In present study euglycemic patients had a better outcome when compared to the admission day hyperglycemic patients. Euglycemic patients had a better recovery after acute stroke. Sixty five percent of euglycemic patients had a good functional recovery. On the contrary, only three percent of admission day hyperglycemic patients had good functional recovery at the end of thirty day follow up. Early inpatient mortality was high in the admission day hyperglycemic patients. Fifty percent of the admission day hyperglycemic patients died within the first thirty days. In the euglycemic patients the early case fatality rate was only fifteen percent. Hence, there was a threefold increased risk of early mortality in the admission day hyperglycemic patients when compared to euglycemics. Poor outcome was noticed in 33.3% of the admission day hyperglycemic patients and in four percent of euglycemic patients.

The present study shows that the admission day elevated blood glucose level was associated with a high early mortality rate and an increased risk of poor functional recovery.

Several studies have confirmed that stress hyperglycemia is associated with a poor outcome (26,27,28). Jorgensen H et al (28), in their large prospective Danish study found that plasma glucose level >11 mmol/L (>198 mg/dL) was associated with a hospital mortality of 17% for non-diabetic patients, 24% for those with known diabetes, and 32% for patients with hyperglycemia with no history of previous diabetes.

In the ischemic stroke group early mortality rate was 9.43 % in euglycemic patients and 56.25% in hyperglycemic patients. Poor outcome was noticed in 5.66 % in euglycemics and 12.5% in hyperglycemics. Hence, hyperglycemia was associated with an increased early mortality rate and poor functional outcome in ischemic stroke group.

In the hemorrhagic patients the early mortality in hyperglycemic patients was 20% and 40 % in euglycemic patients.

In the non-diabetes ischemic stroke patients stress hyperglycemia had worst outcome when compared to euglycemic group. The early mortality rate was 29.7 % in stress hyperglycemics and 9.43% in euglycemics. Hence, this study shows a three-and-a-half-fold increased risk of mortality in non-diabetes stress hyperglycemic patients when compared to non-diabetic euglycemic patients.

The study by Singh KG et al revealed a higher mortality with stress hyperglycaemics and diabetic groups, which was also consistent with other series as reported by some authors (29,30). However, some other studies did not show any significant increase in diabetic stroke deaths in diabetics compared to non-diabetics (31).

In our study group of 158, patients with NIHSS score <5 had good prognosis n=51 (2.51±1.94), patients with score between 5–15 had moderate prognosis n=34(10.41± 2.24), patients with score between 15 to 20 had had poor prognosis n=30(18.23±1.45), patients score more than 20 died within 3 days of admission n=43(0.000). Therefore, as score increases severity of outcome increased.

Similar result found in Gajurel BP et al (32), showing better outcome with mean NIHSS score of 4.6±2.2, and poor outcome with mean NIHSS score of 14.16±7.96.

According to Lindsberg PJ and Roine RO (33), hyperglycemia found in two thirds (66%) of all ischemic stroke patients. In our study, hyperglycemia was noticed in 56% of patients in general and in 53% of patients with ischemic stroke. In Lindsberg PJ and Roine RO study, known diabetes and newly diagnosed diabetes contributed to one third of cases (33%). In our study the same group contributed to 28%.

A study by Christensen H and Boysen G (34) demonstrated that an elevated glucose level after acute stroke was associated with higher stroke severity than those with normal level. The mean NIHSS was 9.5 in euglycemics and 17.3 in hyperglycemic patients.

A study by Umpierrez GE et al (35) confirmed that patients with newly detected hyperglycemia had a significant higher early mortality and a lower functional outcome than patients with a history of diabetes or normoglycemia.

Capes SE et al in their metaanalysis concluded that hyperglycemic patients had threefold increased early mortality than euglycemic patients. After ischemic stroke admission hyperglycemia was associated with threefold increased 30 day mortality than euglcemics. After hemorrhagic stroke, admission hyperglycemia was not associated with higher mortality in either diabetic or non-diabetic patients. In our study, in ischemic patients who had elevated admission day glucose level experienced a three and a half fold increased early mortality than euglycemics.

The study clearly shows an increased early mortality rate and poor functional recovery in patients with diabetes and stress hyperglycemia when compared to euglycemics. Hence, there is an urgent need to confirm the improvement in these patients by normalizing blood sugar. Several trails are now under way to improve the outcome of Stroke by normalizing the blood glucose with human recombinant insulin.

Vinychuk SM et al (36) showed that administration of insulin to patients with hyperglycemia improves the functional recovery and vital activity of mild to moderate ischemic stroke patients. However, other clinical benefits of the insulin therapy remain to be determined.

Studies on larger number of patients and investigating the effect of thrombolytic therapy and lowering hyperglycemia on stroke outcome were recommended.

To improve the outcome, intensive glucose lowering therapies should be tried for a minimum of 72 hrs after acute ischemic stroke.