Risk Factors for Stroke and Resultant Response Time for Victims in by hcj


									     Risk Factors for Stroke and Resultant Response Time for Victims in the State of Georgia:

                                         Subset of a Pilot Study

Marjory Givens*, Mila Prill*, Linda Schieb*, Anne Michal Stevens*

Biostatistics (500-Elon) Laboratory, Department of Biostatistics, Rollins School of Public Health,

Emory University, 1615 Clifton Rd., Atlanta, GA 30322, USA

* Authors contributed equally

                                                BIOS 500
                                              Final Project

       Strokes can result in a debilitating loss of motor skills and cognitive abilities. Thus, it is important

to understand the risk factors involved in order to minimize the number of victims. This study utilized a

set of data from a pilot study for a stroke registry in the state of Georgia in order to investigate potential

associations between age or total lipid count and the onset of stroke as well as to explore a possible

association between race and response time for stroke victims. Our findings indicate a significantly

younger average age of onset for hemorrhagic stroke compared to ischemic stroke (p = 0.0128). In

assessing the potential association between serum cholesterol and stroke, the observed average total lipid

level among the patients in this data set was not significantly different from the recommended cholesterol

level and was below the national average for adults in the United States. Moreover, total lipid levels did

not appear to differ significantly for individuals with a history of smoking compared to individuals with

no history. Finally, there was no significant difference in time between onset of symptoms and arrival at

hospital for African Americans and Caucasians.


       Strokes occur when blood vessels supplying blood to the brain hemorrhage or become blocked by

a blood clot. The consequences of these events are extremely destructive, resulting in damage to the

surrounding nerve cells and cell death within a few hours. Unfortunately, for a victim of stroke, this

translates into impaired brain function and body control.

       Several potential risk factors for stroke have been defined, including age, race, gender, high

cholesterol, and history of smoking. Age is considered a risk factor for stroke, particularly ischemic and

hemorrhagic stroke. Ischemic stroke is caused by blockage in an artery that supplies blood to the brain by

a blood clot, resulting in a deficiency of blood flow (ischemia). Hemorrhagic stroke is caused by bleeding

of ruptured blood vessels (hemorrhage) in the brain. Past studies have shown an increased risk of stroke

among younger populations.1,2 Indeed, one study indicated that the average age for hemorrhagic stroke

victims was less than that of ischemic stroke victims. 2 Thus, further clarification of this risk factor is


       Although previous studies have failed to find a significant association between total lipid levels

overall and occurrence of stroke3-5 , these studies did show that the associations differed by sex3 , and

varied between different types of strokes. The National Average of total cholesterol in US adults is

currently estimated to be 203 mg/dL and only 19% of US adults are estimated to have total lipid levels

above 240 mg/dL.6 Therefore, understanding the relative average total lipid count for patients with stroke

compared to the American Heart Association recommended value of 200 mg/dL is of importance.

       Several studies have found a positive correlation between smoking and cholesterol levels. 7-9 A

history of smoking might lead to the onset of stroke through an increase in cholesterol levels resulting in

the hardening of arteries, restricted blood flow, coronary heart disease and the development of blood clots

in the arteries.10 Investigating lipid counts among the patients who reported a history of smoking

compared to those who did not could provide added insight.

       The duration of time that passes between the onset of symptoms and arrival at a hospital is

clinically relevant, since optimizing treatment may depend on how soon treatment is administered. The

critical windows of time vary from about 1 ½ to 12 hours depending upon the particular treatment. 11-13

Several prior studies have indicated that African Americans suffering from onset of stoke take a

significantly longer amount of time to arrive at a hospital than Caucasians. 14,12 One study suggested that

this difference exists because African American patients were more likely than Caucasians to use their

own source of transportation rather than an ambulance to travel to the hospital. 12 In other studies, no

meaningful difference in arrival time was found between different racial groups seeking treatment for

incident stroke symptoms.15,16 Nevertheless, it is important to uncover potential differences in response

time for these groups in order to improve the quality of care for stroke victims.

       This study is intended to explore these potential risk factors in the Georgia community in the

hopes of improving care for stroke victims. We will examine age variations between the different types of
stroke, average total lipid counts in stroke victims, variations in average total lipid count by history of

smoking, and differences in response time by race.


       The data set obtained includes a subset of the total data acquired by the Coverdell/CDC Georgia

Stroke Registry Pilot Program. This data was collected for strokes occurring between December 1, 2001

and February 28, 2002, from a random sample of 33 hospitals in the state of Georgia. The data set

analyzed herein included 507 patients collected within one year of the study mentioned above. Variables

investigated are shown in Table I.

       To carry out a series of hypothesis tests, we utilized SASv8 software from The SAS Institute,

Cary, North Carolina. A statistically significant difference was considered to be a p-value less than 0.05,

consistent with standards arbitrarily set in the literature. For the hypothesis that the mean age at onset for

hemorrhagic stroke patients would be less than the mean age at onset for ischemic stroke patients we

compared the ages of ischemic stroke patients with the ages of hemorrhagic stroke patients.               We

examined the data and determined that the sample sizes (N=307 and N=87) would allow us to assume

normality for the sample distribution by the central limit theorem.         Based on this assumption, we

performed a 2-sample, independent, T-test to decide whether the hemorrhagic stroke patients' ages were

statistical significantly lower than the ischemic stroke patients' ages. Although the F-test indicated that

the variances were very similar, we chose to report the Satterthwaite (unpooled) T-test results since they

are more conservative.

       A 1-sample T-test was performed to assess if the average total lipid levels for our patient

population were significantly different from a chosen null value of 200 mg/dL. In addition, to address the

hypothesis that the average total lipid levels for our patient population was significantly greater in

individuals who had a history of smoking compared to those who had no history of smoking a 2-sample

T-test (one-tailed) was performed and the Satterthwaite (unpooled) statistic reported.        To investigate

whether or not African Americans suffering from the onset of stroke symptoms took more time to arrive

at the hospital than Caucasians, we removed four observations that had negative values for the amount of

time from onset to arrival since negative times did not make sense in the analysis of incident cases of

stroke. Although the data did not appear to be very normally distributed, we felt that the sample sizes

(167 African American, 313 Caucasians) were sufficient to justify using a T-test for two independent

samples. Due to the dissimilar sample sizes and the large difference in varia nces for the values between

the two groups, we again report the Satterthwaite (unpooled) statistics. In addition, we repeated the

analysis after removing four probable outliers in the data indicating greater than 8 days in duration from

onset to arrival that we feared could be unreasonably skewing our results. The results of the T-tests were

consistent, so we chose to leave those four observations in the analysis.


       Our results show that hemorrhagic stroke patients were, on average, significantly younger than

ischemic stroke patients (p-value of 0.0128, Table 1). No significant increase was found in the average

total lipid level compared to the recommended cholesterol level. The average was 200.69 (95% CI

192.48-208.9) mg/dL, which has a p-value of 0.4344 (Table 1). Furthermore, our data showed no

significant difference between observed total serum lipid levels among individuals with a history of

smoking versus those with no history of smoking (p-value = 0.2595) (Table 1).

       The results of our analysis regarding the time from onset of symptoms to arrival at hospital

between African American and Caucasian patients indicate that there was not a significant increase in

duration for African Americans in this study (p-value = .0639) (Table 1).


       Our results support past research showing that hemorrhagic stroke strikes a younger population

than ischemic stroke.12 This result occurred despite inclusion of a possible outlier (age = 0, a possible

value for a premature infant) in the ischemic stroke group. The recognition of age disparities between the

stroke groups can help to create educational materials and treatments specific to the age groups affected.

       Our analysis of a possible association between total serum lipid levels of our study population and

the occurrence of stroke revealed no significant association. This finding is consistent with other similar

studies3-5 , although our analysis did not explore sex-specific associations. Our statistical results perhaps

could be adjusted for this characteristic of the patient population, but would still be unlikely to

demonstrate significant differences from other studies. Furthermore, this finding actually demonstrates the

lipid total of our patients to be below the national average for US adults. 6

       Our results suggest no association between history of smoking and total serum lipid levels. Total

lipid count is comprised of a measurement of lo w-density lipoproteins (LDL) and high-density

lipoproteins (HDL). The healthier HDL may be the more important for preventing stroke. 5 Interestingly,

low HDL levels associated with smoking in females can apparently be quickly reversed with cessation of

smoking.17 However, we were unable to ascertain the components of the total lipid count or the period of

cessation for those who claimed a history of smoking.

       A stroke registry such as the one being pilot tested would be useful in assessing whether efforts to

educate the public about the importance of seeking prompt treatment for suspected symptoms of stroke

are having any impact. It could also help to identify factors that contribute to delays in receiving

treatments. It would have been interesting investigate the duration of time to treatment after adjusting for

prior medical history of stroke, since one might reasonably hypothesize that those who have previously

suffered from a stroke might recognize the onset of symptoms more quickly and be more aware of the

urgency surrounding seeking treatment for those symptoms.

Table 1. The association of risk factors for stroke among 507 patients admitted to 33
hospitals in the state of Georgia occurring between December 1, 2001 and February
28, 2002.
  Characteristic                    N       Mean         S.D.      Median      Lowest,       p-value
  Age (years)                      507     66.759      15.697      69.000       0; 98       p=0.0128

  Ischemic Stroke                  307     67.762      15.635         70         0; 98
  Hemorrhagic Stroke                87     63.149      17.135         63         7; 94

  Average Total Lipid Count        167     200.689     53.748      194.000     97; 529      p=0.4344

  Average Total Lipid Count        167     200.689     53.748      194.000     97; 529      p=0.2595
  History of Smoking                66     197.394     52.076        189       130; 480
  No History of Smoking            101     202.841     54.962        198       97; 529

  Time From Onset To               503     470.481    2298.181     125.000    5; 45,420     p=0.0639
  Arrival (minutes)
  Caucasian                        313     323.968     813.281     120.000    5; 13,020
  African American                 167     781.347    3,817.769      139      15; 45,420

* The p-value for the average total lipid count represents a comparison to the recommended average total
lipid count.


1.    Bevan H, Sharma K, Bradley W. Stroke in young adults. Stroke 1990;21(3):382-6.

2.    Mettinger KL, Soderstrom CE, Allander E. Epidemiology of acute cerebrovascular disease before
      the age of 55 in the Stockholm County 1973-77: I. Incidence and mortality rates. Stroke

3.    Bots ML, Elwood PC, Nikitin Y, Salonen JT, Freire de Concalves A, Inzitari D, Sivenius J,
      Benetou V, Tuomilehto J, Koudstaal PJ, Grobbee DE. Total and HDL cholesterol and risk of
      stroke. EUROSTROKE: a collaborative study among research centres in Europe. J Epidemiol
      Community Health 2002;56 Suppl 1:i19-24.

4.    Konishi M, Iso H, Komachi Y, Iida M, Shimamoto T, Jacobs DR, Jr., Terao A, Baba S, Sankai T,
      Ito M. Associations of serum total cholesterol, different types of stroke, and stenosis distribution
      of cerebral arteries. The Akita Pathology Study. Stroke 1993;24(7):954-64.

5.    Lindenstrom E, Boysen G, Nyboe J. Influence of total cholesterol, high density lipoprotein
      cholesterol, and triglycerides on risk of cerebrovascular disease: the Copenhagen City Heart
      Study. Bmj 1994;309(6946):11-5.

6.    National Heart L, and Blood Institute. National Cholesterol Education Program: Program
      Description. http://www.nhlbi.nih.gov/about/ncep/ncep_pd.htm 2004.

7.    Khurana M, Sharma D, Khandelwal PD. Lipid profile in smokers and tobacco chewers--a
      comparative study. J Assoc Physicians India 2000;48(9):895-7.

8.    Billimoria JD, Pozner H, Metselaar B, Best FW, James DC. Effect of cigarette smoking on lipids,
      lipoproteins, blood coagulation, fibrinolysis and cellular components of human blood.
      Atherosclerosis 1975;21(1):61-76.

9.    Razay G, Heaton KW. Smoking habits and lipoproteins in British women. Qjm 1995;88(7):503-8.

10.   Horenstein RB, Smith DE, Mosca L. Cholesterol predicts stroke mortality in the Women's Pooling
      Project. Stroke 2002;33(7):1863-8.

11.   Kay R, Woo J, Poon WS. Hospital arrival time after onset of stroke. J Neurol Neurosurg
      Psychiatry 1992;55(10):973-4.

12.   Lacy CR, Suh DC, Bueno M, Kostis JB. Delay in presentation and evaluation for acute stroke:
      Stroke Time Registry for Outcomes Knowledge and Epidemiology (S.T.R.O.K.E.). Stroke

13.   Srivastava AK, Prasad K. A study of factors delaying hospital arrival of patients with acute stroke.
      Neurol India 2001;49(3):272-6.

14.   Kothari R, Jauch E, Broderick J, Brott T, Sauerbeck L, Khoury J, Liu T. Acute stroke: delays to
      presentation and emergency department evaluation. Ann Emerg Med 1999;33(1):3-8.

15.   Barsan WG, Brott TG, Broderick JP, Haley EC, Levy DE, Marler JR. Time of hospital
      presentation in patients with acute stroke. Arch Intern Med 1993;153(22):2558-61.

16.   Morris DL, Rosamond W, Madden K, Schultz C, Hamilton S. Prehospital and emergency
      department delays after acute stroke: the Genentech Stroke Presentation Survey. Stroke

17.   Moffatt RJ. Effects of cessation of smoking on serum lipids and high density lipoprotein-
      cholesterol. Atherosclerosis 1988;74(1-2):85-9.


To top