Keywords
Diabetes - hemorrhagic cerebrovascular events - hyperglycemia - hypertension - ischemic
cerebrovascular events - outcomes
Introduction
Diabetes is an established, independent risk factor for the development of stroke.[[1]] Diabetes mellitus poses about four times higher risk for stroke.[[2]] The mechanisms that contribute to this increased risk include direct vascular injury
due to oxidative stress and a pro-thrombotic state associated with hyperglycemia.[[3]] Other risk factors such as hypertension, dyslipidemia, and obesity coexist in patients
with diabetes frequently.[[2]]
The prevalence of diabetes in the Gulf region is reported to be among the highest
globally, estimated at 19% in the adult population.[[4]] A recent systematic review on stroke epidemiology in the Middle East suggested
an increase in stroke cases over the last decade.[[5]] However, there is limited information on the prevalence of strokes in the Gulf
region or its relationship to glucose intolerance. Hence, we aimed to examine this
relationship using the data from a busy tertiary center in the UAE.
Patients and Methods
Study design
This was a retrospective review of the patient's electronic medical records from January
to December 2019 to examine the relationship between stroke outcomes and dysglycemia.
Objectives
We aimed to (1) determine the percentage of stroke types in patients with and without
diabetes, (2) determine the impact of hyperglycemia on the length of hospital stay
(LOS), (3) determine the effect of hyperglycemia on in-hospital mortality, (4) determine
the incidence of stroke recurrence at 3 and 6 months in patients with and without
hyperglycemia, and (5) determine the incidence of 30-day stroke readmission in patients
with hyperglycemia.
Study population
We included patients with and without diabetes presenting with cerebrovascular events,
both ischemic and hemorrhagic, in a tertiary care center in the UAE. We defined stroke
based on clinical focal neurological signs confirmed by radiological imaging studies.
We included all patients admitted with a clinical and radiological confirmed diagnosis
of stroke (both ischemic and hemorrhagic under the care of medical teams). We excluded
patients admitted with stroke who were eligible for thrombolytic therapy or required
surgical intervention. Those patients were managed by the neurologists and neurosurgeons,
respectively, in dedicated stroke units. The selection of patients for the study is
highlighted in [[Figure 1]], and the number of patients whose follow-up data was available at 1, 3, and 6 months
is shown in [[Figure 2]].
Figure 1: Patient selection flow chart
Figure 2: Follow-up of patients for outcomes
Data variables and outcomes
Hemoglobin A1c (HbA1c) was used as a marker of dysglycemia to categorize patients
into three groups based on the American Diabetes Association criteria. These were
classed as normal: Hba1c <5.7%, prediabetes: Hba1c 5.7%–6.4%, and diabetes: Hba1c:
Equal to or above 6.5%.
We recorded demographic details and other contributory risk factors such as hypertension,
history of ischemic heart disease, and smoking. The duration of hospital stay was
calculated as the time from admission until discharge from the hospital. The adverse
outcomes recorded include in-hospital mortality during the index admission, the incidence
of stroke recurrence at 6 months and 1 year, and the incidence of readmission due
to any other medical cause in the same group of patients in 30 days from the date
of onset of the stroke.
Statistical analysis
Statistical analyses were carried out using the IBM SPSS Statistics for Windows, version
20.0 (IBM Corp, Armonk, NY, USA). The Kolmogorov–Smirnov test was used to determine
the distribution of data. The mean ± standard deviation (SD) was described for variables
with normal distribution and median (interquartile range [IQR]) for variables with
skewed distribution among the continuous data. Descriptive statistics were used to
summarize the data, which was presented as numbers (percentages), means (SD), as appropriate.
As age, body mass index (BMI), and HBa1c follow a normal distribution, a comparative
analysis was carried out using the Analysis of variance (Parametric) under the assumption
of mean and SD.
LOS is a time factor and does not follow a normal distribution. Hence, a comparative
analysis of LOS was carried out by the Kruskal–Wallis test (Non-parametric ANOVA)
under the assumption of mean and IQR.
All categorical data were summarized by frequency and percentage, whereas the Chi-square
test made comparison among groups. The outcome variables were compared by Fisher's
exact test, as the standard Chi-square test is valid only if the cell frequency is
above 5 years.
Results
A total of 496 patients were eligible for the inclusion in the study. They were subdivided
into three main categories based on their glycemic status according to the Hba1c values
at presentation: Nondiabetes (190), prediabetes (117), and diabetes (189). The mean
age of whole cohort was 57 years, with a higher proportion of males (n = 356) than
females (n = 140). There were proportionally more ischemic strokes (n = 350) than
hemorrhagic strokes (n = 147).
A higher proportion of female patients, 42 (35.9%) were in the prediabetes group and
the highest number of males, 143 (75.7%), in the diabetes category, as shown in [[Table 1]].
Table 1: Patient demographics and baseline characteristics
Hypertension was found to be the most prevalent coexisting risk factor among the cohort.
74.2% of all stroke patients had hypertension.
Smoking was evident in 22.2% of the cohort, with equal proportions among the three
categories. 15% of the entire cohort had ischemic heart disease, with a higher proportion
among those with diabetes. On average, the BMI was 27 across the three categories
of patients.
The mean LOS was 5 days for all patients, irrespective of their glycemic status as
shown in [[Figure 3]]. We subdivided the patients into two groups regarding the duration of hospital
stay, those who stayed for 7 days or less and those who stayed for more than 7 days.
Among all the three categories of patients, as shown in [[Table 2]] and [[Figure 2]], more than 50% of patients were discharged in <7 days.
Figure 3: Hospital course according to glycemic status
Table 2: Hospital course according to glycemic status
In-hospital mortality was 4%, with slightly higher mortality (5.3%) in the group with
diabetes; however, this was not statistically significant. The glycemic status did
not seem to impact the stroke readmiss'ion at 3 or 6 months significantly. Furthermore,
the proportion of patients readmitted within 30 days of discharge was not different
between the three categories as highlighted in [[Table 3]].
Table 3: Outcomes according to glycemic status
Discussion
Diabetes is a multi-system disorder characterized by both microvascular and macrovascular
complications. Cardiovascular and cerebrovascular diseases continue to pose a particular
burden in patients with diabetes despite advances in diagnostic and therapeutic interventions.
This concept is not a novel idea; however, when it comes to reporting outcomes and
utilizing prognostic scales, the literature is fraught with considerable controversy
as no single consensus protocol is available to unify the reporting of results.
Several organizations have attempted to define and standardize the reporting of clinical
outcome measures for patients with strokes. Drozdowska et al. in a recent review of
prognostic scales for acute stroke, summarizes the various problems encountered in
trying to identify the most accurate methods of reporting predictive rankings in patients
with strokes.[[6]]
Several studies have shown an association between comorbid diabetes and increased
mortality, LOS, readmission rates, and worse functional and rehabilitation outcomes
after stroke.[[7]] In contrast, other studies have reported no significant differences in poststroke
outcomes between people with or without diabetes.[[8]]
Our study included a cohort of patients admitted for stroke management to a general
medical unit. They were not eligible for thrombolytic therapy at the time of presentation
and did not require any surgical interventions. As per our hospital policy, only those
eligible for thrombolytic treatment or who require surgical intervention receive care
in a dedicated stroke unit. Internists manage the rest in the general medical units.
The proportion of ischemic strokes (71%) was higher than those presenting with hemorrhagic
events (29%) irrespective of the glycemic status. This observation appears to be a
global phenomenon where ischemic events occur more frequently than hemorrhagic events.
Ischemia is the most common stroke type, comprising 85% of all stroke presentations
in the developed world.[[9]],[[10]] El-Hajj et al. found that ischemic stroke was the most frequent type in the Middle
East, followed by intracerebral hemorrhage and subarachnoid haemorrhage.[[5]]
We found a sex imbalance with a higher proportion of events, both ischemic and hemorrhagic,
in males (72%) than females (28%). This observation is similar to other studies within
our region, reporting 75% occurrence in males when compared to females.[[5]]
The number of expatriate patients, 433 (87.3%), far exceeded the number of UAE nationals,
63 (12.7%), in keeping with the country's population demographics. In general, the
UAE population comprises approximately 80% expatriates and only 20% UAE nationals.
Our cohort's mean age was 57 years, slightly younger than what is reported in neighboring
countries. Most patients presented in the sixth and seventh decade in a recent review
on stroke outcomes in the Middle East.[[5]] In the same study, hypertension was the most common reported risk factor, followed
by diabetes. In our cohort, 74% of the patients were hypertensive and 15% were found
to have underlying ischemic heart disease.
BMI and smoking are other relevant risk factors for cerebrovascular events. In our
study, the mean BMI was 27, and 22% of the cohort were identified as active smokers.
El-Hajj et al. reported a case fatality rate of 12%–32% within 1 month of stroke occurrence.[[5]] We found an in-hospital mortality rate of 4% in our cohort. Hyperglycemia impacted
stroke outcomes negatively in patients with diabetes for both ischemic and hemorrhagic
stroke types 9.[[9]],[[10]] However, studies that have examined the effects of glucose-lowering in the setting
of acute strokes have shown variable results.[[11]] In a prospective survey by Tziomalos, stress hyperglycemia did not appear to be
directly associated with the outcome of acute ischemic stroke.[[12]] However, a more recent review by Li et al., with over 8000 participants, reported
that stress hyperglycemia, measured by glucose/HbA1c ratio, was associated with increased
risk of severe neurological deficit and mortality within 1 year in the people with
acute ischemic stroke.[[13]] We found that glycemic status did not seem to impact stroke readmission at 6 or
12 months significantly. Furthermore, the proportion of patients readmitted within
30 days of discharge was not different between the three categories based on their
glycemic status.
Ying-Sung proposed that the predictive power of glycemic measures for poor neurological
outcomes did not differ significantly between the normoglycemic patients and those
with diabetes.[[14]] In a more recent publication, Guo et al. suggest that stress hyperglycemia may
have a higher risk of stroke recurrence than previously diagnosed diabetes.[[15]]
We recognize some limitations of our study. First, it was a retrospective review in
a single center with many patients lost to follow-up due to repatriation to their
home countries (most patients were expatriates) and follow-up in other health care
facilities. We did not report on the NIHSS stroke severity at presentation, which
is an essential prognostic indicator of outcomes. We did not collect the data on other
risk factors such as dyslipidemia or the presence of atrial fibrillation. Hyperglycemia
was based on Hba1c measurements and not complemented by the presenting glucose levels
on admission. Hba1c is well known to be unreliable in certain conditions, and this
adds to the limitations. As it was a retrospective chart review, we could not ascertain
the exact stroke etiology in all cases. Due to these limitations, the results of the
study cannot be generalized to all stroke patients.
Conclusions
The relationship between hyperglycemia and stroke outcomes remains intriguing. Numerous
studies conducted in different regions report variable and conflicting results. This
might be explained by the studies' design, populations, and most importantly, perhaps
the tools and measures used for reporting outcomes. This study adds considerable knowledge
on the epidemiology of strokes in the UAE and is the first large observational study
to report on the relationship between dysglycemia and stroke outcomes.
Our region needs extensive prospective studies that can better elucidate the intricate
relationship between hyperglycemia and stroke outcomes. Diabetes remains a significant
public health problem in our area, with rising numbers projected over the coming years.
Increasing public awareness of the association between diabetes and strokes is essential
to prevent the devastating consequences of strokes.
Authors' Contributions
All authors contributed to the conception of the study, data collection and analysis,
and drafting and revision of the manuscript. They all approved the final version of
the manuscript.
Compliance with ethical principles
The study was conducted ethically as per the World Medical Association Declaration
of Helsinki. The study protocol was approved by DSREC (Dubai Scientific Research Ethics
Committee).Code Number: DSREC-02/2020_12.
Data availability:
Data is available by reasonable requests to the corresponding authors.
