Peripheral glucose levels and cognitive outcome after ischemic stroke

Abstract

Introduction

The relationship between glucose metabolism and stroke outcome is likely to be complex. We examined whether there is a linear or non-linear relationship between glucose measures in the acute phase of stroke and post-stroke cognition, and whether altered glucose metabolism at different time intervals (long- and short-term before stroke, acute phase) is associated with cognitive outcome.

Patients and methods

In all, 664 consecutively recruited patients with acute ischemic stroke and without pre-stroke dementia were included in this prospective observational study. Blood samples were taken at admission and fasting on the first morning after stroke. Duration of diabetes was assessed by interview. Cognitive outcome was assessed by the Telephone Interview for Cognitive Status 3 months post-stroke. Dose-response analyses were used to investigate non-linearity. Regression analyses were stratified by diabetes status and adjusted for relevant confounders.

Results

Cognitive status was testable in 422 patients (81 with diabetes). There was a non-linear relationship between both admission and fasting glucose levels and cognitive outcome. Lower glucose values were significantly associated with lower Telephone Interview for Cognitive Status scores 3 months post-stroke in patients without diabetes with a similar trend in diabetic patients. There was an inverse association between duration of diabetes and Telephone Interview for Cognitive Status scores (linear regression: −0.10 (95% confidence interval: −0.17 to −0.02) per year increase of diabetes duration), whereas HbA_1c was not related to cognitive outcome. Results were supported by sensitivity analyses accounting for attrition.

Conclusion

Lower glucose levels in the acute phase of stroke are associated with worse cognitive outcome but the relationship is non-linear. Long-term abnormalities in glucose metabolism are also related to poor outcome but this is not the case for shorter term abnormalities. Altered glucose levels at different stages of stroke may affect stroke outcome through different pathways.

Keywords

Glucose diabetes cognitive outcome acute ischemic stroke prospective study

Introduction

Altered glucose metabolism is common in patients with acute ischemic stroke¹ and related to poor functional outcome. It is therefore evaluated as a potential target for interventions.² Hyperglycemia in the acute phase is associated with poor functional outcome.^3–9 However, its impact on cognitive outcome is less clear.^10–12 In addition to acute hyperglycemia, chronic hyperglycemia, in the context of (pre)diabetes, also is a risk factor for poor functional outcome after stroke^7,13 and has been shown to be associated with post-stroke dementia.¹⁴ However, little is known about the impact of chronic hyperglycemia on post-stroke cognition as a continuous variable assessed through cognitive scales.

The relationship between glucose metabolism and stroke outcome is complex. For one, there is some indication that in addition to high glucose levels, low glucose levels also negatively influence functional outcome¹⁵ thus raising the possibility of a non-linear relationship between glucose levels and outcome. Second, altered glucose levels at different stages of stroke (e.g. long- or short-term before stroke, upon admission or in the subacute phase) may differently affect outcome and also act through different mechanisms. Chronic hyperglycemia causes both micro- and macrovascular damage^1,16,17 and affects brain structure,¹⁸ whereas acute hyperglycemia can aggravate ischemic damage via acute metabolic effects that act through disturbed recanalization and by increasing reperfusion injury.^1,19 Third, some measures of glucose metabolism may in part reflect secondary effects, an example being admission glucose, which is influenced by stress response.¹

In the current study, we explored the effects of different measures of altered glucose metabolism on cognitive outcome after ischemic stroke. Specifically, we examined whether there is a linear or non-linear relationship between glucose measures in the acute phase and cognitive outcome 3 months after stroke as assessed by cognitive testing. We further investigated the effects of altered glucose metabolism at different time intervals in relation to stroke on cognitive outcome. We used admission glucose and fasting glucose the first day after admission as proxies for glucose metabolism in the hyperacute and subacute phase, duration of diabetes as a proxy for chronic hyperglycemia before stroke, and HbA_1c as an intermediate measure capturing the last weeks prior to stroke.

Methods

Study design and study population

Patients were drawn from the “Munich Stroke Cohort” (MSC), an ongoing observational hospital-based cohort study conducted at the Klinikum der Universität München (KUM), a tertiary level hospital at Ludwig-Maximilians-University. MSC started enrollment in February 2011 and recruits patients of both sexes aged ≥18 years who are admitted to the central stroke unit at KUM. The criterion for admission is a suspected TIA or stroke.

Inclusion criterion for the MSC is a final diagnosis of ischemic or hemorrhagic stroke as defined by an acute focal neurological deficit in combination typical imaging findings. For ischemic stroke, this is a diffusion weighted imaging (DWI)-positive lesion on MR imaging, or a new lesion on a delayed CT scan. Exclusion criteria for the MSC are: (a) time since symptom onset >7 days or unknown, (b) conditions interfering with follow-up such as end-stage malignancies, missing language skills, living far away, hearing or vision impairment, or alcohol abuse; (c) critical medical conditions precluding approaching the patient and family member for study reasons, and (d) inability to consent and no informant available.

For the current study, we restricted the analyses to patients with an acute ischemic stroke and no evidence for dementia before stroke onset as evidenced by a prior diagnosis of dementia according to ICD-10 criteria or a summed score >64 in the short version of the “Informant Questionnaire on Cognitive Decline in the Elderly” (IQCODE). We further restricted the analysis to patients with valid Telephone Interview for Cognitive Status (TICS) scores 3 months post-stroke. Enrollment for the current study occurred between February 2011 and May 2015. A flow-chart detailing the study profile is presented in Supplement Figure 1.

Standard protocol approvals, registrations, and patient consent

The study was conducted according to the Declaration of Helsinki and approved by the local ethics committee. Written informed consent was obtained by the patient or legal guardian prior to study participation. Ethical approval for surrogate consent was obtained to minimize recruitment bias.

Data collection

Information on demographic variables, living situation, pre-stroke functional and cognitive status (pre-stroke modified Rankin Scale, IQCODE), lifestyle habits, health history as well as medication before stroke was provided by the patient or the next of kin at baseline examination. Stroke severity was assessed using the National Institutes of Health Stroke Scale (NIHSS) by a certified study clinician. Stroke etiology was determined according to the criteria of the TOAST classification. A detailed description of the study protocol of the MSC is outlined in the Supplement.

Exposure measurement

Information on diagnosis of pre-stroke diabetes was provided at baseline examination and months and year of diagnosis of diabetes was recorded. Blood samples were taken at admission and between 6 and 9 a.m. on the first morning following admission. Admission and fasting glucose and HbA_1c values were obtained from the central laboratory as part of routine diagnostic blood draws. Diabetes was defined as patient’s self-report of diabetes or use of antidiabetic drugs or HbA_1c ≥ 6.5%.¹

Outcome measurement

Cognition was assessed 3 months post-stroke by the German version of the TICS, which is composed of 11 items with scores ranging from 0 to 41 (higher scores indicate better performance).^20–22 TICS reflects a validated test for global cognitive function after stroke and can equally be performed in telephone and face to face visits.^21–23

Statistical analysis

Analyses were stratified for patients with and without diabetes. To check for linear associations between measures of glucose and cognitive function, we used dose-response analyses based on restricted cubic spline functions (SAS macro RCS_Reg V1.3 beta) and created graphs with knots at 5th, 25th, 50th, 75th, and 95th percentiles of glucose levels.²⁴ The analysis provides the Wald test for non-linear association and overall association. A significant result for non-linear association means that the association between exposure and outcome is significantly not linear. If a clear peak could be identified between the 5th and 95th percentiles, the respective value (exposure) was used as reference. Linear regression was performed if there was no evidence of non-linear association. The basic model for all analyses was adjusted for age, sex, and education. Models further adjusting for vascular risk factors, pre-stroke cognitive impairment, and glucose measures are defined in the Supplemental Methods and the results are presented in Supplemental Figures 2 to 5.

As impaired cognitive function is associated with untestability²⁵ and increased dropout in longitudinal studies,^26–28 we analyzed the effect of the exposure on the chance to have missing TICS data (Supplemental Figure 6).

We report odds ratios (OR) for logistic regression, effect estimates per unit increase for linear regression, 95% confidence intervals (CI) and p-values. All analyses are two-sided, conducted at a 0.05 level of significance and carried out using SAS version 9.3 (SAS Institute Inc, Cary, NC).

Results

Study population

From 1195 eligible patients, 664 agreed to participate into the study. In 422 patients, cognitive status was testable 3 months post-stroke (median: 88 days, Q1–Q3: 81–99 days) without missing items on the TICS. Reasons for loss to follow up and incomplete assessment are detailed in Supplemental Figure 1. Characteristics of patients with and without cognitive assessments are given in Supplemental Table 1. Patients not cognitively assessed more often had a history of stroke, had more severe stroke, and were functionally and cognitively more impaired both before and after stroke.

Figure 1.

Glucose levels and cognitive outcome 3 months post-stroke. (a) Admission glucose and (b) fasting glucose. Knots on the solid red line mark the 5th, 25th, 50th, 75th, and 95th percentiles of glucose levels. Dotted grey lines indicate the 95% confidence interval. The regression is adjusted for age, sex, and education. As a clear peak could be identified between the 5th and 95th percentiles, the respective value was used as reference and marked by a vertical green dotted line.

Table 1.

Baseline characteristics of the study population stratified by history of diabetes mellitus.

	Patients without diabetes (n = 341; 81%)	Patients with diabetes (n = 81; 19%)	p-value
Female	138 (40)	25 (31)	0.11
Age (years)	70 (61–77)	72 (65–80)	0.035
Pre-stroke living situation
Alone at home	102 (30)	23 (28)	0.93
At home with family/friends	237 (69)	58 (72)
Institution	2 (1)	0
Pre-stroke mRS > 1	37 (11)	7 (9)	0.56
Pre-stroke cognitive decline	54 (16)	18 (22)	0.15
Education ≤12 years	151 (44)	35 (43)	0.86
Smoking status
Never	164 (48)	28 (35)	0.086
Ex-smoker	110 (32)	34 (42)
Current smoker	67 (20)	19 (23)
Physical activity
Low	76 (22)	28 (35)	0.039
Moderate	118 (35)	27 (34)
High	147 (43)	25 (31)
Abdominal obesity	160 (49)	51 (65)	0.009
History of cardiovascular risk factors
Arterial hypertension	210 (62)	62 (77)	0.011
Dyslipidemia	118 (35)	35 (43)	0.15
Previous stroke	32 (9)	14 (17)	0.040
Atrial fibrillation	80 (23)	24 (30)	0.25
Depression	32 (9)	5 (6)	0.36
Blood pressure (mmHg)
Systolic	141 (129–151)	148 (136–157)	0.003
Diastolic	80 (72–88)	79 (72–87)	0.90
Baseline NIHSS	2 (1–4)	3 (1–5)	0.24
Baseline mRS	1 (0–2)	2 (1–2)	0.030
Baseline MoCA	26 (23–28) (n = 302)	26 (22–28) (n = 74)	0.32
TOAST
Large artery atherosclerosis	50 (15)	12 (15)	0.21
cardio-embolic	87 (26)	21 (26)
Small artery occlusion	36 (11)	16 (20)
Other	18 (5)	3 (4)
Competing etiologies/undefined	150 (44)	29 (36)
Laboratory findings
Admission glucose (mg/dl)	109 (99–124) (n = 329)	150 (114–202) (n = 78)	<0.001
Fasting glucose (mg/dl)	96 (89–104) (n = 340)	130 (106–172) (n = 81)	<0.001
HbA_1c (%)	5.6 (5.3–5.8) (n = 310)	6.9 (6.4–8.2) (n = 75)	<0.001
LDL cholesterol (mg/dl)	129 (105–156)	119 (99–149)	0.049
HDL cholesterol (mg/dl)	49 (42–60)	42 (35–52)	<0.001
Triglyceride (mg/dl)	108 (81–145)	137 (109–200)	<0.001
Serum CRP (mg/dl)	0.4 (0.2–0.7)	0.5 (0.2–1.1)	0.064

Data are given as median (quartile (Q) 1–(Q) 3) or number (percentage). Associations between patients without and with diabetes and baseline-characteristics were assessed using Wilcoxon–Mann–Whitney test and χ²-test/Fishers exact test. Significant results are shown in bold.

mRS: modified Rankin Scale; MoCA: Montreal Cognitive Assessment; NIHSS: National Institutes of Health Stroke Scale. For definition of cardiovascular risk factors and pre-stroke cognitive decline, see Supplemental Methods. Information on diagnosis of pre-stroke diabetes was provided by the patient (99%) or the next of kin (1%).

Table 1 shows baseline characteristics of the study population (n = 422) stratified by history of diabetes. Patients with diabetes (n = 81) were older, more often physically inactive, and more often had abdominal obesity, hypertension, and a previous stroke. They had higher values of glucose, HbA_1c, and triglycerides and lower values of HDL and LDL cholesterol.

Admission glucose and cognitive outcome

Admission glucose was missing for 12 patients without diabetes and 3 patients with diabetes. Glucose levels ranged from 65 to 232 mg/dl in patients without diabetes and from 81 to 550 mg/dl in patients with diabetes.

Patients without diabetes

There was a significant overall association between glucose levels and TICS values (p = 0.01) with evidence for a non-linear relationship (Figure 1; p = 0.01). The glucose level with the highest (i.e. best) TICS score (reference point, for definition see methods) was 100 mg/dl. Lower glucose values were associated with lower TICS scores whereas there was no association between glucose values above the reference point and TICS scores (Table 2). However, patients with higher glucose levels were more likely to have missing TICS scores than patients with lower glucose levels (Supplemental Figure 6). The results remained stable using different adjustments (Supplemental Figure 2).

Figure 2.

HbA_1c and cognitive outcome 3 months post-stroke. Patients without diabetes: non-linear association: p = 0.97; linear regression: −0.35 (95% CI: −1.26 to 0.57), p = 0.46; patients with diabetes: non-linear association: p = 0.96; linear regression: −0.24 (95% CI: −0.79 to 0.21), p = 0.29. Knots on the solid red line mark the 5th, 25th, 50th, 75th, and 95th percentiles of HbA_1c levels. Dotted grey lines indicate the 95% CI. The regression is adjusted for age, sex, and education.

Table 2.

Association between admission and fasting glucose and cognitive outcome 3 months post-stroke.

	Reference	5th percentile	25th percentile	50th percentile	75th percentile	95th percentile
Admission glucose
Patients without diabetes	100 mg/dl	88 mg/dl	99 mg/dl	110 mg/dl	126 mg/dl	166 mg/dl
Patients without diabetes		−1.21 (−1.89 to −0.52)	−0.00 (−0.06 to 0.06)	−0.56 (−1.33 to 0.20)	−0.52 (−1.33 to 0.29)	−0.23 (−1.32 to 0.86)
Patients with diabetes	125 mg/dl	95 mg/dl	118 mg/dl	155 mg/dl	200 mg/dl	337 mg/dl
Patients with diabetes		−1.78 (−3.84 to 0.28)	−0.09 (−0.37 to 0.19)	−0.66 (−2.37 to 1.05)	−0.94 (−2.93 to 1.04)	−2.25 (−4.49 to −0.02)
Fasting glucose
Patients without diabetes	92 mg/dl	80 mg/dl	89 mg/dl	97 mg/dl	106 mg/dl	131 mg/dl
Patients without diabetes		−1.42 (−2.20 to −0.64)	−0.08 (−0.41 to 0.24)	−0.13 (−0.61 to 0.36)	−0.31 (−1.01 to 0.39)	0.08 (−0.85 to 1.01)
Patients with diabetes	130 mg/dl	84 mg/dl	106 mg/dl	131 mg/dl	171 mg/dl	231 mg/dl
Patients with diabetes		−1.98 (−4.21 to 0.25)	−0.75 (−2.49 to 0.99)	−0.01 (−0.07 to 0.05)	−2.02 (−3.66 to −0.38)	−2.76 (−4.84 to −0.69)

The reference was the glucose value associated with the highest TICS score; the glucose values at the corresponding percentiles were compared with the reference; estimates are age, gender, and education adjusted differences in TICS scores (95% confidence interval); significant results are shown in bold.

TICS: Telephone Interview for Cognitive Status.

Patients with diabetes

There was a trend for an association between glucose levels and TICS values (Figure 1). However, the association was statistically not significant (non-linear association: p = 0.23; overall association: p = 0.17). The reference point was 125 mg/dl. TICS scores were lower both with lower and higher glucose values (Figure 1, Table 2) and again, patients with higher glucose levels were more likely to have missing TICS scores (Supplemental Figure 6). The results remained stable using different adjustments (Supplemental Figure 2).

Fasting glucose and cognitive outcome

Fasting glucose was available for all but one patient (Table 1). Glucose levels ranged from 62 to 178 mg/dl in patients without diabetes and from 67 to 397 mg/dl in patients with diabetes.

Patients without diabetes

There was a significant overall association between glucose levels and TICS values (p < 0.01) with evidence for a non-linear relationship (Figure 1; p = 0.02). The reference point was 92 mg/dl. Lower glucose values were associated with lower TICS scores, whereas there was no association between glucose values above the reference point and TICS scores (Figure 1, Table 2). Patients with glucose levels below and above the reference point were more likely to have missing TICS scores (Supplemental Figure 6). The results remained stable using different adjustments (Supplemental Figure 3).

Patients with diabetes

There was a trend for an association between glucose levels and TICS values (Figure 1). However, the association was statistically not significant (non-linear association: p = 0.11; overall association: p = 0.08). The reference point was 130 mg/dl. Both lower and higher glucose values were associated with lower TICS scores (Figure 1, Table 2). Sensitivity analyses showed no clear relationship between glucose levels and missing data on the TICS (Supplemental Figure 6). The results remained stable using different adjustments (Supplemental Figure 3).

HbA_1c and cognitive outcome

HbA_1c was missing for 31 patients without diabetes and 6 patients with diabetes. Values ranged from 3.7 to 6.4% in patients without diabetes and from 5.3 to 12% in diabetic patients.

There was no evidence for a non-linear or linear relationship between HbA_1c and cognitive outcome in patients with or without diabetes (Figure 2). Sensitivity analyses showed no clear relationship between HbA_1c levels and missing data on the TICS (Supplemental Figure 6). The results remained stable using different adjustments (Supplemental Figure 4).

History of diabetes, duration of diabetes, and cognitive outcome

There was no significant association between history of diabetes and TICS score (−0.53 (95% CI: −1.26 to 0.20), p = 0.15). However, missing TICS scores were more frequent in patients with than without diabetes (OR = 1.30 (0.88–1.92)). Among patients with diabetes there was a linear inverse association between the duration of diabetes and TICS scores (Wald test for non-linear association: p = 0.95; linear regression: −0.10 (95% CI: −0.17 to −0.02) per year increase of duration, p = 0.015; Figure 3). Moreover, a longer diabetes duration was associated with missing TICS scores (Supplemental Figure 6). The results remained stable using different adjustments; however the association was statistically not significant (Supplemental Figure 5).

Figure 3.

Duration of diabetes and cognitive outcome 3 months post-stroke in diabetics. Knots on the solid red line mark the 25th, 50th, 75th, and 95th percentiles of duration of diabetes. Dotted grey lines indicate the 95% CI. The regression is adjusted for age, sex, and education.

Discussion

The main finding of this study is the non-linearity in the association between glucose levels in the acute phase of ischemic stroke and cognitive outcome. Specifically, we found both higher and lower glucose levels to be associated with poor cognitive outcome. Another novel finding is the inverse association between diabetes duration and cognitive outcome. Our results imply that altered glucose levels at different stages of stroke may affect stroke outcome through different mechanisms.

Although there is quite an extensive literature on the relationship between hyperglycemia in the acute phase of ischemic stroke and outcome, we are not aware of studies that systematically examined the association of lower glucose levels with cognitive outcome. The observed association between lower glucose levels and worse cognitive outcome might be explained by at least two aspects. First, it is conceptually plausible that insufficient glucose supply to tissue that is already experiencing (borderline) energy failure can result in both neuronal loss and permanent disability.^29–31 Of note, low glucose levels occurred in patients with and without diabetes and in the latter group therefore definitely represent a spontaneous event, unrelated to glucose lowering treatment. The association with worse outcome was observed even at glucose levels that do not cause neuroglycopenia under normal circumstances. In people without diabetes, the threshold for neuroglycopenic symptoms is around 54 mg/dl; i.e. 3 mmol/l.^32,33 If it is indeed the case that modest reductions in blood glucose levels cause secondary damage in brain tissue that is already depleted of energy due to ischemia, this could offer a lead for therapy not only to treat high but also low levels of glucose. However, this would need to be tested in trials and these may be complicated in recruitment if, as in our study, only few patients are in really low glucose ranges. An alternative explanation would be that the association might be non-causal. It could be that people in a worse physical state prior to stroke and at risk of poor cognitive function are less likely to be able to increase their glucose levels in response to stress. In this case, lower glucose levels would be an epiphenomenon.

To our knowledge, the current study is the first that considers the possibility of a non-linear relationship between glucose values in the acute phase und cognitive outcome after stroke. Importantly, we found both admission glucose and fasting glucose to display a non-linear relationship with cognitive outcome. The two measures are influenced by different external factors including food intake for admission glucose and treatment related factors for fasting glucose on the first morning. Together, these findings emphasize the strength of the observed association. While preliminary, our data suggest that the optimal glucose values may be different for patients with diabetes compared with patients without diabetes (Figure 1), which may be relevant for therapeutic considerations.¹⁹

While the majority of previous studies showed a negative effect of higher admission glucose on functional outcome after stroke,^3–7 findings for cognitive outcome have been inconsistent both in patients with^10–12,34 and without stroke.¹⁶ In the current study, there was a trend for a negative relationship between glucose levels above the reference point and poor cognitive outcome in patients with diabetes and this was seen both for admission glucose and fasting glucose (Figure 1). In contrast, we found no such trend in patients without diabetes. A possible explanation may be that (a) clinically relevant hyperglycemia was scarce in patients without diabetes, and (b) that patients with higher glucose levels were more likely to have missing TICS scores. Untestability has been shown to be associated with post-event dementia.²⁵ Hence, there may be a bias from attrition.

Extreme glucose values in the acute phase of stroke supposedly act through short-term effects on the injured brain. In contrast, chronic disturbances in glucose metabolism likely act through multiple and mixed effects. Patients with advanced diabetes are at a twofold increased risk of both micro- and macrovascular damage.¹ This is commonly explained by the exposure to chronic hyperglycemia as well as treatment-associated hypoglycemic episodes typically in later disease stages.^35,36 There are surprisingly few studies on the effect of diabetes duration on cognitive function. Our findings on the effect of diabetes duration are in line with previous data from the general population showing that a longer exposure to diabetes is more harmful.^37–39

HbA_1c is considered a measure of the average circulating glucose over the preceding months and hence reflects the metabolic state prior to stroke in our study. A negative effect of higher HbA_1c on functional outcome after ischemic stroke has been shown in some^40,41 but not all⁸ studies. Our results are in line with a previous study on cognitive outcome¹² as we found no statistically significant association between HbA_1c and TICS score 3 months post-stroke. There was no evidence that attrition affected this result. However, a possible effect of HbA_1c on long-term cognitive outcome (i.e. 12 months after stroke and beyond) remains to be investigated.

Strengths of this study include the prospective design, the assessment of multiple measures of glucose metabolism considering different stages of stroke, the examination of non-linearity in glucose measures, the use of cognitive testing, and sensitivity analyses accounting for missing data. The main limitation of this study was the small sample size in the subgroup with diabetes. The overall association between glucose levels and cognitive outcome was not significant in patients with diabetes. However, effect estimates for lower glucose levels were even larger for patients with diabetes compared with patients without diabetes. Also, variations in duration of diabetes may have resulted in a mixture of different mechanisms mediating cognitive outcome and this in turn may have influenced association measures. Additional limitations are a potential bias from the recruitment of less severely affected patients and residual confounding. Of note, however, our results remained stable throughout all models using different adjustments (Supplemental Tables 2 to 5). Our findings are further supported by sensitivity analyses, thus accounting for potential bias from selective attrition.

Conclusion

This study shows that lower glucose values in the acute phase of stroke have a negative impact on stroke outcome, even glucose values normally not considered dangerously low.²⁹ We further found longstanding diabetes but not HbA_1c to be associated with cognitive outcome 3 months post-stroke thus supporting the hypothesis of differential mechanisms of long-term diabetes prior to stroke and extremes in admission or fasting glucose values in the acute phase of stroke on the brain. The complexity in the relationship between glucose metabolism and cognitive outcome may in part explain why there is currently no evidence for a beneficial effect of glucose-lowering treatment on outcome after stroke.^1,42,43

Supplemental Material

Supplemental material for Peripheral glucose levels and cognitive outcome after ischemic stroke—Results from the Munich Stroke Cohort

Supplemental Material for Peripheral glucose levels and cognitive outcome after ischemic stroke—Results from the Munich Stroke Cohort by Vera Zietemann, Frank Arne Wollenweber, Anna Bayer-Karpinska, Geert Jan Biessels and Martin Dichgans in European Stroke Journal

Footnotes

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: GJB consults for and receives research support from Boehringer Ingelheim and consults for Takeda Pharmaceuticals. Compensation for these services is transferred to his employer, the UMC Utrecht.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Deutsche Forschungsgemeinschaft (German Research Foundation) within the framework of the Munich Cluster for Systems Neurology (EXC 1010 SyNergy) and the Vascular Dementia Research Foundation.

Ethical approval

The study was conducted according to the Declaration of Helsinki and approved by the local ethics committee (project number 366-10).

Informed consent

Written informed consent was obtained by the patient or legal guardian prior to study participation.

Guarantor

MD.

Contributorship

VZ and MD conceived the study, developed the protocol, and gained the ethical approval. FAW and ABK were involved in patient recruitment and acquisition of data. VZ performed the data analysis and VZ, MD, and GJB interpreted the data and wrote the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

Acknowledgements

The authors thank Margit Deschner, Angelika Dörr, Adelgunde Zollver, Christoph Gerlach (study nurses), Rupert Aigner, Timo Hasselwander (medical documentalist), and Mareike Schmitt (student research assistant) for their support.

References

Luitse

Biessels

Rutten

et al.

Diabetes, hyperglycaemia, and acute ischaemic stroke. Lancet Neurol 2012; 11: 261–271.

Gray

Hildreth

Sandercock

et al.

Glucose-potassium-insulin infusions in the management of post-stroke hyperglycaemia: The UK Glucose Insulin in Stroke Trial (GIST-UK). Lancet Neurol 2007; 6: 397–406.

Luitse

van Seeters

Horsch

et al.

Admission hyperglycaemia and cerebral perfusion deficits in acute ischaemic stroke. Cerebrovasc Dis 2013; 35: 163–167.

Hsieh

Chen

et al.

The prognostic roles of initial glucose level and functional outcomes in patients with ischemic stroke: Difference between diabetic and nondiabetic patients. Disabil Rehabil 2012; 34: 34–39.

Capes

Hunt

Malmberg

et al.

Stress hyperglycemia and prognosis of stroke in nondiabetic and diabetic patients: A systematic overview. Stroke 2001; 32: 2426–2432.

Fuentes

Ortega-Casarrubios

Sanjose

et al.

Persistent hyperglycemia >155 mg/dl in acute ischemic stroke patients: How well are we correcting it?: Implications for outcome. Stroke 2010; 41: 2362–2365.

Desilles

Meseguer

Labreuche

et al.

Diabetes mellitus, admission glucose, and outcomes after stroke thrombolysis: A registry and systematic review. Stroke 2013; 44: 1915–1923.

Cao

Ling

et al.

Higher fasting glucose next day after intravenous thrombolysis is independently associated with poor outcome in acute ischemic stroke. J Stroke Cerebrovasc Dis 2015; 24: 100–103.

Kim

Jahan

Saver

et al.

Impact of glucose on outcomes in patients treated with mechanical thrombectomy: A post hoc analysis of the solitaire flow restoration with the intention for thrombectomy study. Stroke 2016; 47: 120–127.

10.

Kruyt

Nys

van der Worp

et al.

Hyperglycemia and cognitive outcome after ischemic stroke. J Neurol Sci 2008; 270: 141–147.

11.

Nys

van Zandvoort

de Kort

et al.

The prognostic value of domain-specific cognitive abilities in acute first-ever stroke. Neurology 2005; 64: 821–827.

12.

Kandiah

Wiryasaputra

Narasimhalu

et al.

Frontal subcortical ischemia is crucial for post stroke cognitive impairment. J Neurol Sci 2011; 309: 92–95.

13.

Newman

Bang

Hussain

et al.

Association of diabetes, homocysteine, and HDL with cognition and disability after stroke. Neurology 2007; 69: 2054–2062.

14.

Pendlebury

Rothwell

. Prevalence, incidence, and factors associated with pre-stroke and post-stroke dementia: A systematic review and meta-analysis. Lancet Neurol 2009; 8: 1006–1018.

15.

Ntaios

Egli

Faouzi

et al.

J-shaped association between serum glucose and functional outcome in acute ischemic stroke. Stroke 2010; 41: 2366–2370.

16.

Geijselaers

Sep

Stehouwer

et al.

Glucose regulation, cognition, and brain mri in type 2 diabetes: A systematic review. Lancet Diabetes Endocrinol 2015; 3: 75–89.

17.

Singh-Manoux

Schmidt

. Diabetes: A risk factor for cognitive impairment and dementia? Neurology 2015; 84: 2300–2301.

18.

Biessels

Reijmer

. Brain changes underlying cognitive dysfunction in diabetes: What can we learn from MRI? Diabetes 2014; 63: 2244–2252.

19.

Wan Sulaiman

Hashim

Che Abdullah

et al.

Managing post stroke hyperglycaemia: Moderate glycaemic control is better? An update. EXCLI J 2014; 13: 825–833.

20.

Brandt

Spencer

Folstein

. The telephone interview for cognitive status. Neuropsychiatr Neuropsychol Behav Neurol 1988; 1: 111–117.

21.

Desmond

Tatemichi

Hanzawa

. The telephone interview for cognitive status (tics): Reliability and validity in a stroke sample. Int J Geriatr Psychiatr 1994; 9: 803–807.

22.

Barber

Stott

. Validity of the telephone interview for cognitive status (tics) in post-stroke subjects. Int J Geriatr Psychiatry 2004; 19: 75–79.

23.

Pendlebury

Welch

Cuthbertson

et al.

Telephone assessment of cognition after transient ischemic attack and stroke: Modified telephone interview of cognitive status and telephone montreal cognitive assessment versus face-to-face montreal cognitive assessment and neuropsychological battery. Stroke 2013; 44: 227–229.

24.

Desquilbet

Mariotti

. Dose-response analyses using restricted cubic spline functions in public health research. Stat Med 2010; 29: 1037–1057.

25.

Pendlebury

Klaus

Thomson

et al.

Methodological factors in determining risk of dementia after transient ischemic attack and stroke: (iii) applicability of cognitive tests. Stroke 2015; 46: 3067–3073.

26.

Glymour

Chene

Tzourio

et al.

Brain mri markers and dropout in a longitudinal study of cognitive aging: The three-city dijon study. Neurology 2012; 79: 1340–1348.

27.

Chatfield

Brayne

Matthews

. A systematic literature review of attrition between waves in longitudinal studies in the elderly shows a consistent pattern of dropout between differing studies. J Clin Epidemiol 2005; 58: 13–19.

28.

Euser

Schram

Hofman

et al.

Measuring cognitive function with age: The influence of selection by health and survival. Epidemiology 2008; 19: 440–447.

29.

Robbins

Swanson

. Opposing effects of glucose on stroke and reperfusion injury: Acidosis, oxidative stress, and energy metabolism. Stroke 2014; 45: 1881–1886.

30.

Whitmer

Karter

Yaffe

et al.

Hypoglycemic episodes and risk of dementia in older patients with type 2 diabetes mellitus. JAMA 2009; 301: 1565–1572.

31.

Wright

Frier

. Vascular disease and diabetes: Is hypoglycaemia an aggravating factor? Diabetes Metab Res Rev 2008; 24: 353–363.

32.

McAulay

Deary

Frier

. Symptoms of hypoglycaemia in people with diabetes. Diabet Med 2001; 18: 690–705.

33.

Frier

. Defining hypoglycaemia: What level has clinical relevance? Diabetologia 2009; 52: 31–34.

34.

Mellon

Brewer

Hall

et al.

Cognitive impairment six months after ischaemic stroke: A profile from the aspire-s study. BMC Neurol 2015; 15: 31.

35.

Biessels

. Sweet memories: 20 years of progress in research on cognitive functioning in diabetes. Eur J Pharmacol 2013; 719: 153–160.

36.

Seaquist

. The impact of diabetes on cerebral structure and function. Psychosom Med 2015; 77: 616–621.

37.

Roberts

Geda

Knopman

et al.

Association of duration and severity of diabetes mellitus with mild cognitive impairment. Arch Neurol 2008; 65: 1066–1073.

38.

Tuligenga

Dugravot

Tabak

et al.

Midlife type 2 diabetes and poor glycaemic control as risk factors for cognitive decline in early old age: A post-hoc analysis of the whitehall ii cohort study. Lancet Diabetes Endocrinol 2014; 2: 228–235.

39.

Rawlings

Sharrett

Schneider

et al.

Diabetes in midlife and cognitive change over 20 years: A cohort study. Ann Intern Med 2014; 161: 785–793.

40.

Hjalmarsson

Manhem

Bokemark

et al.

The role of prestroke glycemic control on severity and outcome of acute ischemic stroke. Stroke Res Treat 2014; 2014: 694569.

41.

Kuwashiro

Sugimori

Ago

et al.

The impact of predisposing factors on long-term outcome after stroke in diabetic patients: The fukuoka stroke registry. Eur J Neurol 2013; 20: 921–927.

42.

Matz

Teuschl

Firlinger

et al.

Multidomain lifestyle interventions for the prevention of cognitive decline after ischemic stroke: Randomized trial. Stroke 2015; 46: 2874–2880.

43.

Bellolio

Gilmore

Ganti

. Insulin for glycaemic control in acute ischaemic stroke. Cochrane Database Syst Rev 2014; 1: CD005346.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.11 MB

Peripheral glucose levels and cognitive outcome after ischemic stroke—Results from the Munich Stroke Cohort

Abstract

Introduction

Patients and methods

Results

Conclusion

Keywords

Introduction

Methods

Study design and study population

Standard protocol approvals, registrations, and patient consent

Data collection

Exposure measurement

Outcome measurement

Statistical analysis

Results

Study population

Admission glucose and cognitive outcome

Patients without diabetes

Patients with diabetes

Fasting glucose and cognitive outcome

Patients without diabetes

Patients with diabetes

HbA1c and cognitive outcome

History of diabetes, duration of diabetes, and cognitive outcome

Discussion

Conclusion

Supplemental Material

Supplemental material for Peripheral glucose levels and cognitive outcome after ischemic stroke—Results from the Munich Stroke Cohort

Footnotes

Declaration of conflicting interests

Funding

Ethical approval

Informed consent

Guarantor

Contributorship

Acknowledgements

References

Supplementary Material

HbA_1c and cognitive outcome