The onset of complete ischemic stroke is associated with significant hemorrhagic disturbance, impairing the blood fluidity comprising of plasma viscosity and erythrocyte aggregation, hence enhancing the progression of ischemia by virtue of their detrimental influence on cerebral blood flow [1]. Erythrocyte aggregation has been proposed as a useful marker to detect inflammatory state in unstable angina [2, 3] and stroke [4], whose evaluation is available to clinician indirectly through erythrocyte sedimentation rate (ESR). Several other inflammatory markers having high specificity and low sensitivity including C-reactive protein and fibrinogen are also available, but the evaluation of these markers is expensive. In contrast, ESR is easy to do; economical and readily available and still remains useful tool for monitoring disease activity in inflammatory conditions as well as an important prognostic factor in non-inflammatory conditions. Elevated ESR count within 72 hours of ischemic stroke has been known to be associated with large brain infarcts [5]. But, this non-specific indicator of inflammation is widely influenced by sex, advanced age, anemia, fibrinogen, extreme obesity and hypercholesterolemia [6]. In this study, we have observed the determinants of ESR and correlated ESR values with various cerebrovascular risk factors involved in ischemic stroke.

The study involved 163 ischemic stroke in patients admitted in Uppal Neuro hospital, Amritsar. The selection of cases was done on the basis of magnetic resonance imaging (MRI) reports or computerized tomography (CT) reports if MRI report was not available. According to WHO 1988, [7] stroke is defined as rapidly developing clinical signs of focal (global) disturbance of cerebral function, lasting more than 24 hours or leading to death with no apparent cause other than of vascular origin. After inclusion, informed consent was taken from each patient and demographic profile was recorded on a pre-prepared questionnaire including age, sex, occupation, dietary habits, habitat, smoking and alcohol consumption. A complete medical history was also recorded for various risk factors like hypertension (HTN), diabetes, coronary artery disease (CAD), atrial fibrillation (AF), migraine and previous history of any cerebrovascular event. Blood pressure was also monitored immediately after admission. Within 24 hours of admission to hospital, blood investigations like hemoglobin (Hb), total leukocyte count (TLC), differential leukocyte count (DLC), red blood cell (RBC) count, ESR, packed cell volume (PCV), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), random blood glucose and serum cholesterol were carried out. Patients having transient ischemic attacks and any autoimmune disease were excluded from study.

1. Laboratory Procedure

Venous blood samples were collected immediately after admission and divided into vacutainers with EDTA for analysis of hematological parameters and into clot activator for serum analysis of cholesterol and random blood glucose with automated blood analyzer (Sysmex -21X, Sysmex Co., Kobe, Japan) and biochemical analyzer (Biochem Erba, Sri Medi Tech, Madurai, India), respectively. ESR measurement (in mm/hr) was done according to Westergren method using EDTA blood diluted with 3.8% sodium citrate. The Wester-gren's method measures the rate of gravitational settling of anticoagulated RBCs in 1 hour from a fixed point in a calibrated tube of a defined length and diameter held in an upright position. At the end of 1 hour, the distance from the meniscus to the top of the column of erythrocytes was recorded as ESR value in units of millimeters per hour.

2. Statistical analysis

Statistical analysis was performed using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). Continuous variables were expressed as mean±standard deviation and evaluated by student t-test to examine the gender differences. Nominal variables were expressed in the form of the number of cases and percentages (%) and evaluated by Pearson chi-square test or Fisher’s exact chi-square test. To determine confounding factors contributing to the variability of ESR, we performed linear regression analysis with ESR as a dependent variable and investigated parameters including age, Hb, TLC, neutrophils, lymphocytes, platelet count, RBC count, PCV, MCV, MCH, MCHC, serum blood glucose and serum cholesterol as an independent variable. To study the effect of age, the patients having age range 41∼90 years were divided into 5 groups and one-way ANOVA was performed to determine the significant difference between these age groups with respect to ESR. The results of the data analyses were considered significant at two-tailed P-value>0.05.

We analyzed the data for 163 ischemic stroke patients: 107 (65.64%) males and 56 (34.36%) females with mean age of 64.82±11.5 and 66.41±12.4 years respectively. The preliminary analysis of demographic features, various clinical risk factors and other relevant laboratory values like Hb, ESR, RBC count, TLC, neutrophils, lymphocytes, PCV, MCV, MCH MCHC, serum cholesterol, systolic and diastolic blood pressure are shown in Table 1. Most of the parameters except ESR and PCV fall in the normal reference range. The mean ESR (17.98±14.21) of patients was slightly higher and mean PCV (38.44±5.1) was slightly lower than normal range. Similarly, blood pressure measurements depicted minor variation (SBP=149.40±22.44, DBP= 88.38±12.04). The distribution of haematological, biochemical parameters and cerebrovascular risk factors stratified on gender basis are presented in Table 2. To evaluate sex variation, we performed student t-test and observed statistically significant differences between two sexes with respect to Hb (t=4.78, P<0.001), ESR (t=29.97, P<0.01), RBC count (t=2.47, P<0.05), PCV (t=3.855, P<0.001), MCH (t=2.43, P<0.05) and MCHC (t=2.06, P<0.05). The respective cardiovascular risk factors depicted no significant difference between two sexes, but diet pattern between males and females revealed highly significant difference showing that more males were on non-vegetarian food (χ²=25.05, P<.001).

Table 1

Demographic and clinical variables characteristics of study population

Variable name (Units)	Total (N=163) Mean±SD/number (%)	Normal Range
Males	107 (65.6%)	-
Females	56 (34.4%)	-
Age (years)	65.37±11.86	-
Haemoglobin (gm%)	12.78±1.87	12∼16 gm%
ESR (mm/hr)	17.98±14.21	0∼15 mm/hr
RBC count (million/cmm)	4.43±0.68	4.3∼5.6 million/cmm
Total leucocyte count	9324.85±3046.94	4,500-11,000
Neutrophils	73.96±11.08	50∼75%
Lymphocytes	22.75±12.01	20∼45%
PCV (%)	38.44±5.1	39∼60%
MCV (fl)	86.28±6.88	78∼98 fl
MCH (pg)	28.21±3.08	26∼33 pg
MCHC (%)	32.65±1.78	30∼36%
Serum Cholesterol (mg%)	176.76±38.47	150∼260 mg%
Blood Glucose	159.22±73.94	Up to 190
SBP (mmHg)	149.4±22.44	140 mmHg
DBP (mmHg)	88.38±12.04	90 mmHg
Hypertension (%)	154 (94.5%)	-
Diabetes (%)	57 (34.97%)	-
CHD (%)	50 (30.7%)	-
AF (%)	7 (4.3%)	-
Migraine (%)	5 (3.1%)	-
TPA treated (%)	13 (7.98%)	-
Previous history of stroke (%)	23 (14.1)%)	-
Diet pattern (%)	Veg=81 (49.7%) Non-veg=82 (50.3%)	-
Alcoholics (%)	56 (34.4%)	-
Habitat (%)	Rural=121 (74.2%) Urban=35 (21.5%) Suburban=7 (4.3%)	-

Abbreviations: TPA, tissue plasminogen activator; ESR, erythrocyte sedimentation rate; RBC, red blood cell; PCV, packed cell volume; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; SBP, systolic blood pressure; DBP, diastolic blood pressure; CHD, coronary heart disease; AF, atrial fibrillation.

All values are means±standard deviation for continuous variables and percent for nominal variables.

Table 2

Gender – wise distribution of haematological, biochemical parameters and cerebrovascular risk factors of ischemic stroke patients

Variable name (Units)	Males (N=107) Mean±SD	Females (N=56) Mean±SD	P-value
Age (years)	64.82±11.5	66.41±12.4	0.209*
Hemoglobin (gm%)	13.3±1.9	11.9±1.42	<.001*
ESR (mm/hr)	15.6±13.7	22.5±14.2	.0032*
RBC (million/cmm)	4.53±0.739	4.25±0.572	.014*
Total leucocyte count	9216±2874	9533±3370	0.529*
Neutrophils	74.3±10.7	73.4±11.9	0.647*
Lymphocytes	22.6±12.1	23.1±11.9	0.777*
PCV (%)	39.5±5.08	36.4±4.51	<.001*
MCV (fl)	86.5±7.12	85.9±6.44	0.603*
MCH (pg)	28.6±3.15	27.4±2.78	.016*
MCHC (%)	32.9±1.81	32.3±1.68	.041*
Serum cholesterol (mg%)	175±40	180±35.4	0.485*
SBP (mmHg)	148±22.9	151±21.6	0.458*
DBP (mmHg)	87.5±11.1	90±13.6	0.215*
Hypertension (N,%)	101 (94.4%)	54 (96.4%)	0.716†
Diabetes (N,%)	39 (36.4%)	18 (32.1%)	0.584†
CHD (N,%)	37 (34.6%)	13 (23.2%)	0.135†
AF (N, %)	3 (2.8%)	4 (7.14%)	0.234†
Migraine (N, %)	2 (1.9%)	3 (5.4%)	0.34†
TPA treated (N, %)	9 (8.4%)	4 (7.1%)	1.00†
Previous history of stroke (N, %)	13 (12.2%)	10 (17.9%)	0.32†
Diet pattern (N, %)	Veg=38 (35.5%) Non-veg=69 (64.5%)	Veg=43 (76.8%) Non-veg=13 (23.2%)	<.001†

Abbreviations: ESR, erythrocyte sedimentation rate; RBC, red blood cell; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; TPA, tissue plasminogen activator.

All values are means±standard deviation (range) for continuous variables and percent for nominal variables.

*One way ANOVA; ^†Pearson Chi square test were P<0.05.

On performing linear regression, we observed significant inverse correlation of Hb, RBC and PCV with ESR for all cases (Table 3). Age did not show a strong positive correlation with ESR. On analyzing the results gender-wise, sex (females) came out to be an important factor in defining values that influence ESR levels. Hb and PCV best explained the variations in ESR values in females as comparison to males. In both sexes, RBC count showed a weak correlation with ESR. Age did not enter the model of regression analysis for men, but in females, both age and serum cholesterol showed weak positive correlation with ESR.

Table 3

Linear regression analysis for erythrocyte sedimentation rate in all cases, males and females

Variable	Pearson correlation	β (S.E)	P-value
All cases (N=163)
Hemoglobin	−.391	−2.971 (.551)	<.001
RBC	−.291	−5.911 (1.532)	<.001
PCV	−.420	−1.167 (.199)	<.001
Age	−.198	.236 (.092)	.011
Males (N=107)
Hemoglobin	−.275	−1.980 (.676)	.004
RBC	−.232	−4.291 (1.757)	.016
PCV	−.296	−.796 (.251)	.002
Age
Females (N=56)
Hemoglobin	−.530	−5.309 (1.155)	<.001
RBC	−.331	−8.226 (3.195)	.013
PCV	−.544	−1.718 (.360)	<.001
Age	.338	.386 (.146)	.011
Serum Cholesterol	−.271	−.109 (.053)	.043

Abbreviation: See Table 1,2.

To study the importance of age, we further divided patients into 5 age groups and evaluated the data by one-way ANOVA but we did not find any significant age-related increase in ESR count.

To speculate the relationship of ESR count with various cardiovascular risk factors, we categorized the patients into three groups according to ESR values as shown in Table 4. Patients with ESR≤10 mm/hr were included in the group I (N=35) with mean age of 61.26±11.42. The higher proportion of cases were included in group II (N=103) having ESR values between 11 and 25 and mean age of 65.96±12.13. Group III included only 25 patients having higher ESR values (ESR>25) and mean age 68.68±10.04. It is evident from results that elder people have elevated levels of ESR, as patients having ESR values above 25 have higher mean age as compared to group I and II, the difference being statistically significant (F=3.05, P<0.05) among three groups. The patients in group III with elevated ESR were more likely to be females (64%) as compared to group II (32.03%) and group I (20%). In contrast to group I and II, these patients also had a higher prevalence of previous history of ischemic stroke (20%), hypertension (96%), diabetes (44%), migraine (8%), higher mean lymphocyte values (25.28±2.16) and low mean levels of Hb (11.36± 1.87), TLC (8972±3139.2), neutrophils (71.2±10.78) RBC count (3.98± 0.62) and PCV (34.16±5.10). It was observed that Hb (F=12.48, P<0.001), RBC count (F=8.24, P<0.001) and PCV (F=15.29, P<0.001) reflected highly statistically significant association with increased ESR values. The results revealed no statistical difference among three groups with respect to TLC, neutrophils, lymphocytes, MCV, MCH, MCHC, SBP and DBP. In terms of ischemic stroke risk factors, hyper-tension, diabetes, migraine was found to be more prevalent in group III, but no statistically significant difference was observed among three groups.

Table 4

Demographic characteristic of ischemic stroke patients according to ESR values

	Group I (N=35)	Group II (N=103)	Group III (N=25)	P-value
Total No. of cases	35	103	25
No. of males	28 (80%)	70(67.9%)	9 (36%)
No. of females	7 (20%)‡||	33 (32%)‡§	16 (64%)‡§||	.001‡
Mean age	61.26±11.42‡	65.96±12.13‡	68.68±10.03‡	<.05*
Previous h/o stroke	4 (11.4%)	13 (12.6%)	5 (20%)	0.534†
Hypertension	32 (91.4%)	99 (96.1%)	24 (96%)	0.508†
Diabetes	13 (37.1%)	33 (32.03%)	11 (44%)	0.507†
CAD	10 (28.6%)	34 (33%)	6 (24%)	0.651†
AF	1 (2.9%)	5 (4.9%)	1 (4%)	1.0†
Migraine	2 (5.7%)	1 (0.9%)	2 (8%)	0.075†
SBP	144±25.57	152.28±21.69	145.12±19.42	0.098*
DBP	86.71±12.89	89.76±11.12	85.04±13.91	0.139*
Hemoglobin	13.64±1.746‡	12.83±1.715‡	11.36±1.871‡	<.001*
RBC	4.69±0.923‡	4.45±0.572‡	3.98±0.616‡	<.001*
PCV	40.91‡	38.63‡	34.16‡	<.001*

Abbreviation: See Table 1,2.

*One way ANOVA.

^†Pearson Chi square test.

^‡The difference among the group I, group II and group III is statistically significant.

^§The difference between the group II and group III is statistically significant (P=0.0065).

^||The difference between the group I and group III is statistically significant (P=0.0014).

More than 70 years after its introduction into clinical medicine by Westergren [8], ESR is still in use as a simple and quick laboratory marker to assess chronic inflammation. Atherosclerosis, which is an important cause of ischemic stroke, is basically low-grade inflammation identified by ESR and C reactive protein levels may be an additional risk factor for development of ischemic stroke or transient ischemic attacks. This readily available time-honoured routine analysis may highlight patients with ongoing thrombosis or fibrinolysis who are at increased risk of further pathological vascular events. Elevated ESR values have been reported to be associated with large ischemic lesions and more severe deficits [9].

Our study documented determinants of ESR values in a group of ischemic stroke patients that included age, sex, Hb level, RBC count, packed cell volume and serum cholesterol (females) that is in line with previously published studies [10, 11]. In our study, patients showed high mean ESR and lower mean haematocrit or PCV than the normal range, but with respect to Hb, only females were found to be anaemic having decreased Hb levels. Increased ESR levels have been detected in patients immediately after ischemic stroke that indirectly reflects the association between degree of acute phase response and extent of local brain damage [12]. Moreover, decreasing and low levels of Hb and Hct have been reported as strong predictor of poor outcome and mortality after ischemic stroke [13].

This study also provides insight into significant gender differences detected in Hb, RBC, PCV, MCH and MCHC as supported by various studies [14-16]. While analysing the determinants of ESR, we observed a significant inverse correlation of Hb, RBC count and PCV with ESR. We observed slightly raised ESR in anaemic patients (Tables 3, 4). An inverse correlation between Hb and ESR count has been demonstrated that may be due to fact that in anaemia, the velocity of the upward flow of plasma is altered so that red blood cell aggregates fall faster [17]. Available evidence in animal models of ischemic stroke suggests that lower Hb reduces the threshold for ischemia and results in large infarct volumes and moreover lower Hb value may adversely affect the energy balance with in penumbra [18]. The haematocrit is the proportion of blood volume occupied by red blood cells and is one of the factors affecting erythrocyte sedimentation. In our study, PCV comes out to be a highly significant determinant, showing strong inverse correlation with ESR. Previous studies have also demonstrated the similar trend between ESR and PCV [19].

Age was not strongly correlated with ESR in our study. Similar studies have been published with respect to age related increase in ESR levels [20, 21]. Also in healthy volunteers, ESR levels have been reported to rise with age. It increases by 0.85 mm/hr for each 5-year increase in age [22, 23]. We noted weak correlation of age with ESR, but on comparing age groups, no significant difference was observed among 5 age groups. This finding that age may not be a significant determinant of ESR is further supported by studies of Feher and colleagues carried on haemorrhagic parameters and aging [24]. They concluded that the correlation of hemorrhagic parameters like fibrinogen, haematocrit, RBC aggregation, plasma and whole blood viscosity and advancing age was inconsistent and these parameters did not correlate with advancing age at all.

An association between ESR values and any risk factor was not seen in our study, yet group III patients showed increased prevalence of HTN, diabetes and migraine as compared to group II and I (Table 4). These findings of our study are supported by previous work of Comoglu and colleagues [25] and they did not find any correlation between ESR levels and risk factors in ischemic stroke patients except for CAD and valvular heart disease. The variations in markers of low-grade inflammation are associated with genetic and familial influences, age, environnmental stresses, cardiovascular risk factors and vascular or nonvascular diseases. One prospective population study has shown the association of inflammatory markers to be associated with future coronary heart disease [26, 27]. It is feasible that low-grade inflammation results into activation of thrombotic system and then adding to these risk factors may enhance the predisposition of an individual towards cerebrovascular events.

The ESR evaluation is an easy-to-perform, inexpensive and time-honored analysis that can help in early detection of low-grade inflammation and eventual atherosclerotic burden. The measures to reduce inflammation of endothelium and atherosclerotic plaques of blood vessels, in the long run can reduce the incidence of stroke and its consequences in predisposed individuals. The hemorheological abnormalities have been found to persist for a long time after an acute stroke. Hence, the investigation of hematological parameters can be supportive to select the optimal medical treatment in the secondary prevention of stroke and further correction of hemorhelogical disturbances can aid in reducing the recurrent stroke like events.

There is multitude of confounding factors affecting ESR levels. Several studies have shown the significant relation of ESR levels with hemoglobin [28], serum fibrinogen [29] and triglycerides [30] suggesting their independence. In our study, the main determinants of ESR came out to be sex, hemoglobin, RBC count and PCV. To conclude, the elevation in ESR levels should not be employed as a direct indicator of inflammation in ischemic stroke. The routine analysis of ESR levels, if elevated, may highlight the patients with ongoing ischemic stroke events, but along with various determinants including hematological parameters and other strong markers of inflammation like C reactive protein, fibrinogen and triglycerides must be taken into account. Also, the ischemic stroke patients also have a wide risk profile displaying multiple risk factors like HTN, diabetes, CAD, smoking etc, it is possible that they present a pre-existing pro-inflammatory pro-coagulant condition, which may at least partly contribute to pathological shift in ESR values soon after cerebral infarction.

We are very thankful to UPE for financial support to KK, Dr. Ashok Uppal for providing samples and all the participants for their timely support to carry out this study.

None

Kauer K¹, Ph.D.; Kauer A², Ph.D.;Kauer A¹, Professor.